Plant Gene Research

Plant Gene Research
Basic Knowledge and Application
Edited by
E.S.Denni~Canbe"a
B. Hohn, Basel
Th. Hohn, Basel (Managing Editor)
P. J. King, Basel
J. Schell, Kiiln
D. P. S. Verma, Montreal
Springer-Verlag Wien New York

A Genetic Approach
to Plant Biochemistry
Edited by A. D. Blonstein
and P. J. King

Dr. Anne D. Blonstein
Friedrich Miescher-Institut, Basel
Dr. Patrick J. King

Friedrich Miescher-Institut, Basel
With 30 Figures
This work is subject to copyright.

All rights are reserved,
whether the whole or part of the material is concerned,
specifically those of translation, reprinting, re-use of illustrations,
broadcasting, reproduction by photocopying machine or similar means,
and storage in data banks.
1986 by Springer-Vedag/Wien
Sof'tcover reprint of the hardcover 1st edition 1986
Library of Congress Cataloging in Publication Data

A Genetic approach to plant biochemistry.
(Plant gene research)
Includes bibliographies and index.
1. Plant biochemical genetics. I. Blonstein, A. D.
(Anne D.), 1958- . II. King, P. J. (Patrick J.),
1941- III. Series.
QK981.3.G46 1986 581.19'2 86-24865
ISSN 0175-2073
ISBN-13: 978-3-7091-7463-0 e-ISBN-13: 978-3-7091-6989-6
DOl: 10.1007/978-3-7091-6989-6
Preface
Biologists ask how the growth, development and behaviour of organisms

happen, how these processes are co-ordinated and how they are regulated
by the environment. Today the questions are phrased in terms of the genes
involved, their structure and the control of their expression. Mutations
(recognised by a change in phenotype) label genes and can be used to
study gene structure, gene function and the organisation of the genome.
This is "Genetics". Study of phenotypes down to the level of the enzymes
and structural proteins coded for by genes is "Biochemistry". It is self
evident that only by studying phenotype ("Biochemistry") can we do "Ge-
netics" and that "Genetics" (perturbation of the phenotype) is the key to
understanding the "Biochemistry". There can surely be no better argu-
ments for a more holistic approach to biology than the massive output of
knowledge from microbial "Biochemical Genetics" and the more recent
revelations from "Molecular Genetic" studies of development in Droso-
phila.
When one remembers that most of the important conceptual developments
in genetics (the discovery of the nucleus, the hereditary mechanism, cyto-
plasmic inheritance, mutation, the one gene-one enzyme hypothesis, semi-
conservative replication of DNA, heterochromatin, transposable elements)
were discovered by biologists working with plants, it is surprising that the
genetic approach to the study of basic plant properties is so underdevel-
oped when compared to all other biological systems. Because the advan-
tages, the directness and reduced ambiguity, of studying normal functions
by perturbing or eliminating single genes are so obvious when compared to
the circumstantial approach of pure physiology and biochemistry, there
must exist strong reasons why the biochemical genetics of plants is still an
emerging field. Undoubtedly there are aspects of biochemistry or
molecular biology common to several organisms that one would not
choose to study with a complex higher plant. There are not that many effi-
cient cytogenetic systems in plants, like those of maize, Arabidopsis or
tomato, which are at the same time suited for particular biochemistry.
There are also various physical criteria for the isolation of mutants that
must be met: Where traits must be screened for by examination of indi-
vidual plants, the investment of the necessary time and space may be inhib-
iting. The selection techniques applicable to whole plants required for the
rescue of specific mutants may not be available. Lethal nutritional mutants
isolated with ease from Neurospora may be impossible to find given the
complex life cycle of the higher plant.
The "Princes of Serendip" were probably not very far away when studies of
VI Preface
genetic aspects of gibberellins (Chapter 1), abscisic acid (Chapter 2), pho-
tosynthesis (Chapter 3) and endosperm proteins (Chapter 7) got under way.
Some areas of investigation were set off initially by spontaneous variation
in a field and a passing, curious biologist. Others started with the analysis
of randomly generated variation following mutagenesis or a dip into a
local seed collection to look for a useful variant. The rapid success of the
alcohol dehydrogenase systems (Chapter 4) and the growth of nitrate
reductase studies (Chapter 5) are clearly due to powerful and simple
chemical selection systems being available for mutants deficient in the
enzymes, and to the conditional lethality of the traits. There are, unfortu-
nately, very few positive selection systems for deficient mutants. Cell and
protoplast culture offer both a method for conveniently selecting or
screening for mutants amongst very large cell populations and a way of res-
cuing lethal nutritional mutants. Isolation of variation is carried out at the
cell level and thus many mutants may be found that would not otherwise
be recognised by plant selection. After selection, plants may be regenerated
and normal genetics and biochemistry undertaken. Cell culture was central
to the isolation of the majority of nitrate reductase deficient mutants
(Chapter 5). An account of the recent successes in the isolation of plant
auxotrophs via cell and protoplast culture is given in Chapter 10.
The life cycle of the majority of higher plants includes the liberation of
enormous numbers of semi-autonomous haploid gametophytes (pollen
grains). The realisation that many important sporophytic genes are
expressed in pollen grains and that the selection for new alleles may occur
in natural populations at this level has stimulated interest in gametophytic
gene expression (Chapter 9) and the possibility that mutant selection using
pollen could increase the variation available to the breeder and the
biochemist. Plant breeding rather than biochemistry is probably the greater
stimulus to research in nodulation and nitrogen fixation in legumes
(Chapter 6) and in plant/pathogen genes (Chapter 8) but, as the authors
point out, in these and so many other areas of plant breeding there is an
increasing need to understand the mechanisms and the molecular events
involved. The protection and stabilisation of our plant resources is a
worthy target for plant biochemical genetics.
Basel, August 1986 A. D. Blonstein and P. J. King
Contents
Chapter 1 Gibberellin Mutants

J. B. Reid, Hobart, Australia
I. Introduction 1
II. Selection and Identification of Mutants 2
III. GA Synthesis 8
IV. Internode Length 8
A. Synthesis Mutants 8
B. Insensitive Mutants 14
V. Seed Dormancy 17
VI. Flowering and Senescence 18
VII. Site of Action of GA Mutants 22
VIII. Conclusions 25
IX. References 26
Chapter 2 Genetic Aspects of Abscisic Acid

M. Koornneef, Wageningen, The Netherlands
I. Introduction 35
A. Abscisic Acid as a Plant Hormone 35
B. Biosynthesis and Metabolism of ABA 36
C. Genetic Aspects of ABA 36
II. ABA-Deficient Mutants 38
A. Isolation, Genetic and General Phenotypic Characteristics 38
B. Pleiotropic Effects of ABA-Deficient Mutants 41
C. Biochemistry of ABA-Deficient Mutants 41
D. The Use of ABA Mutants in Water Relations Research 44
E. The Use of ABA Mutants in Seed Physiology 44
F. ABA Deficiency in Relation to Other Physiological Effects 46
III. Mutants Affecting ABA Sensitivity 46
IV. Genetic Differences in ABA Accumulation 48
V. Conclusions 50
VI. References 50
VIII Contents
Chapter 3 Mutants as Tools for the Elucidation of Photosynthetic

Processes
C. Critchley and W. Bottomley, Canberra, Australia
I. Introduction 55
II. Two Genomes Code for the Structural ~md Regulatory Elements of
the Photosynthetic Apparatus 59
III. Identification of Thylakoid Membrane Proteins and Their
Functions 61
IV. From Phenotype to Gene Structure 62
A. Herbicide Resistance 62
B. Barley 63
C. Chlamydomonas 64
D. Arabidopsis 65
E. Oenothera 65
F. Summary 66
V. Conclusions 68
VI. References 69
Chapter 4 Maize Alcohol Dehydrogenase: A Molecular Perspective

W. L. Gerlach, M. M. Sachs, D. Llewellyn, E. J. Finnegan
and E. S. Dennis, Canberra, Australia and St. Louis, Mo.,
U. S.A.
I. Introduction 74
II. Genetics and Expression of ADH Enzymes in Maize 74
A. Two Genes Encode AD H in Maize 74
B. Organ Specificities of ADHI and ADH2 Activities 75
C. Mutations of Adh Genes of Maize 75
D. The Maize Anaerobic Response 76
III. Isolation of Adh Genes 78
A. cDNAs from Anaerobically Induced Maize Genes 78
B. Isolation and Identification of Adhl cDNA Clones 79
C. Isolation and Identification of Adh2 cDNA Clones 81
D. Isolation of Adh Genes from Other Plant Species 81
IV. Structure of Plant Adh Genes 82
V. Three-Dimensional Structure of ADH Enzymes 84
VI. Genetic Change Around and Within Adh Genes of Maize 86
A. Allelic Variation 86
B. Somaclonal Variation 87
C. Gene Duplication 87
D. Ds Element Mutations in Adhl 88
E. Robertson's Mutator 88
VII. Approaches to the Mechanism of Adh Gene Regulation 89
A. DNA Sequence Comparisons 89
B. Mutations in Adh Which Mfect Expression 92
VIII. In Vivo Expression of Adh 93
Contents IX
A. Attempts at Expression in Cereal Tissue Culture Cells 94

B. Testing Maize Adh Gene Activity in Dicotyledonous Plant
Cells 95
C. Expression of Maize Adh in Animal Cells 95
IX. Conclusions 96
X. References 96
Chapter 5 The Molecular Genetics of Higher Plant Nitrate Assimilation

J. L. Wray, St. Andrews, Scotland
I. Introduction 101
II. The Nitrate Assimilation Pathway 102
A. Nitrate Uptake 102
B. Nitrate Reductase 103
C. Nitrite Reductase 108
III. Genetics of Nitrate Assimilation 108
A. Introduction 108
B. Nitrate Uptake Mutations 110
C. Nitrate Reductase Mutations 111
D. Nitrite Reductase Mutations 130
E. Regulatory Alterations in Nitrate Assimilation 131
F. Conclusions 133
IV. Applied Aspects 134
A. Somatic Hybridisation 134
B. Cloning Nitrate Assimilation Genes 135
C. Plant Gene Transfer Systems 140
D. Whole Plant Studies 141
V. References 143
Chapter 6 Plant Genetic Approaches to Symbiotic Nodulation and

Nitrogen Fixation in Legumes
P. M. Gresshoff and A. C. Delves, Canberra, Australia
I. Introduction 158
II. A General Description of Legume Nodule Ontogeny 161
III. The Parasponia-Bradyrhizobium Symbiosis 168
IV. Biochemical and Molecular Analysis of Plant Functions 170
V. Gene-for-Gene Aspects of Nodulation 171
VI. Existing Plant Variation in Symbiotic Nitrogen Fixation 174
VII. Existing Single Locus Variation for Nodulation-Nitrogen
Fixation 177
VIII. Induced Mutation in Symbiotic Characters 178
A. Pea and Chickpea Mutants 179
B. Soybean Nodulation Mutants 181
x Contents
IX. Conclusions 194

X. References 195
Chapter 7 Endosperm Proteins

P. I. Payne, Cambridge, U. K.
I. Introduction 207
II. Origin and Development of the Endosperm 208
III. Classification of the Major Endosperm Proteins 209
IV. Biochemical Complexity and Genetic Variation of Endosperm
Proteins 211
V. Gene Mutations 214
A. Role in Plant Breeding 215
B. Role in Genetics 217
C. Role in- Biochemistry and Molecular Biology 218
D. Role in the Food Industry 222
VI. Chromosome Mutations 223
VII. Conclusions 226
VIII. References 227
Chapter 8 Molecular Approaches to Plant and Pathogen Genes

R. I. S. Brettell and A. J. Pryor, Canberra, Australia
I. Introduction 233
II. A Molecular Approach to Gene-for-Gene Resistance 235
A. The Shotgun Method 236
B. Transposon Mutagenesis or Gene Tagging 238
III. The Role of Toxins in Plant Disease 239
IV. Conclusions 242
V. References 243
Chapter 9 Gametophytic Gene Expression

D. L. Mulcahy, Amherst, Mass., U. S. A.
I. Introduction 247
II. Overlap Between Sporophytic and Gametophytic Genotypes 249
III. Gametophytic Gene Expression and the Angiosperms 251
IV. The Influence of Haploid Genotype on Pollen Size 252
V. Time of Gene Expression in Pollen 252
VI. Methods of Haploid Selection 252
VII. Gametophytic Gene Expression and the Style 254
VIII. Gene Expression in the Megagametophyte 255
Contents XI
IX. Conclusions 255

x. References 256
Chapter 10 Auxotroph Isolation In Vitro
A. D. Blonstein, Basel, Switzerland
I. Introduction 259
A. Some Preliminary Comments 259
B. Why Auxotrophs? 260
C. The Characteristics, Advantages and Disadvantages of In vitro
Systems for Mutant Isolation 261
D. Species Suitable for Auxotroph Isolation 262
II. Mutagenesis 263
III. Methods of Selection 264
A. Conditions and Media 264
B. Positive Selection 265
C. Negative Selection 266
IV. Phenotypes 270
A. Amino-Acid, Vitamin and Purine Auxotrophs 270
B. Nitrate Reductase Mutants 271
C. Temperature-Sensitive Mutations 271
V. Fusion and Transformation 272
VI. Conclusions and Future Prospects 273
VII. References 274
Subject Index 281

Chapter 1
Gibberellin Mutants
James B. Reid
Department of Botany, University of Tasmania, Hobart, 7001, Australia
With 3 Figures
Contents
I. Introduction
II. Selection and Identification of Mutants
III. GA Synthesis
IV. Internode Length
A. Synthesis Mutants
B. Insensitive Mutants
V. Seed Dormancy
VI. Flowering and Senescence
VII. Site of Action of GA Mutants
VIII. Conclusions
IX. References
I. Introduction
An enormous amount has been written on the gibberellins (GAs) during

the last decade (for reviews see Graebe and Ropers, 1978; Hedden et al.,
1978; MacMillan, 1980) culminating in the extensive work by Crozier
(1983). During this time considerable advances have been made regarding
their biosynthesis and metabolism due to the use of GC-MS techniques
and the production of both radio- and stable-isotope labelled compounds
(see Hedden, 1983). However, information is still lacking on the develop-
mental processes controlled by endogenous GAs, the site of action of the
GAs and the mechanism(s) by which GAs elicit a physiological response.
Indeed it is still debated if endogenous GA levels are responsible for con-
trolling developmental processes (Trewavas, 1981; Ingram et al., 1983;
Phinney, 1984) and which of the over sixty identified GAs are biologically
active (e. g. Hoad, 1983; Phinney, 1984).
A range of mutants in the fungus Gibberella fujikuroi has been used in
2 James B. Reid
biosynthetic studies of the GAs. For example, mutant BI-41 a, which blocks
GA biosynthesis prior to ent-kaurenoic acid, has been used extensively
since the virtual absence of endogenous GAs allows the use of unlabelled
compounds in metabolism studies (see Bearder, 1983). Unfortunately,
similar use has not been made of GA synthesis mutants to help elucidate
the GA biosynthetic pathways in higher plants. However, GA synthesis
mutants in higher plants do provide some of the strongest evidence
available to show which developmental processes are under the control of
endogenous GAs and which of the multitude of GAs are biologically
active. They are also the basis for many of the most widely used bioassays
for GA-like substances. For example, the dwarf mutants dj in maize, dx in
rice and Ie in peas have provided useful information on endogenous levels
of biologically active GAs in many physiological studies. Mutants may also
be useful as probes for examining the site and mode of action of the GAs,
especially as this has proven difficult using other techniques (Stoddart,
1983).
The present article concentrates on the mutants which are available in
higher plants, how they have been used to show which developmental
responses are controlled by endogenous GAs and how mutants may be
used in the future to determine the mode and site of action of the GAs. It
has not been possible to cover all areas in which mutants have been used in
the study of GAs in higher plants. For example, the work using GA syn-
thesis mutants as bioassays, and the extensive metabolic work in which
often non-native radio-labelled compounds have been fed to mutants such
as dj maize or dwarf (Ie) peas have not been examined in detail. Detailed
reviews of this work can be found elsewhere (e. g. Graebe and Ropers,
1978; Sponsel, 1983a).
II. Selection and Identification of Mutants
The selection of mutants influencing GA metabolism or sensitivity has

usually been based upon phenotypic changes in characteristics associated
with the GAs in application experiments. These characters include
internode length (e. g. Brian and Hemming, 1955; Phinney, 1956), apical
senescence (Davies et al., 1977; Proebsting et al., 1978), sex expression
(Shifriss, 1973) and seed dormancy (Koomneef and van der Veen, 1980). A
list of proposed GA mutants is included in Table 1. Many of these mutants
arose spontaneously and have subsequently been selected for research or
agricultural purposes. For example the Ie gene in peas confers the dwarf
habit and this characteristic was mentioned at least as early as 1597 by
Gerarde (see Blixt, 1972). It was presumably selected because it reduced
lodging of the crop and is today incorporated in many green pea varieties.
Other mutants have been produced by either radiation or chemical
mutagens (e. g. Sidorova, 1970; Wellensiek, 1969; Koomneef and van der
Veen, 1980). Unfortunately the selection of mutants involves the growth of
large numbers of plants and no efficient self-selection techniques are
Gibberellin Mutants 3
available. However, selection based upon characters evident early in the

life-cycle may be beneficial. For example, Koornneef and van der Veen
(1980) have shown that the more extreme GA synthesis mutants of Arabid-
opsis thaliana L. have increased dormancy in addition to a dwarf habit.
Table 1. The phenotypic effect and proposed mode of action of certain genes sug-
gested to influence GA-synthesis or sensitivity
Species Gene Phenotype Proposed Gene Key Reference(s)

Action
Zeamays d} dwarf blocks GA20 to GAl Spray et al., 1984

d2 dwarf blocks C-7 -oxidation Phinney and Spray,
1982
d3 dwarf blocks C-13-hydrox- Phinney and Spray,
ylation 1982
d5 dwarf blocks B activity of Hedden and Phinney,
ent-kaurene syn- 1979
thetase
anI strongly blocks GA synthesis Phinney, 1961
male, dwarf
nal dwarf GA-insensitive Phinney, 1961
na2 dwarf GA-insensitive Phinney, 1961
D8 dwarf GA-insensitive Phinney, 1961
pel dwarf GA-insensitive Phinney, 1961
mi2 dwarf GA-insensitive Phinney, 1961
Pisum sativum Ie dwarf blocks GA20 to GAl Ingram et al., 1984
la cry' slender saturates GA Potts et aI., 1985
receptors
la cry" crypto partially saturates GA Potts et al., 1985
receptors
Ik erectoides GA insensitive Reid and Potts, 1986
Ih dwarf blocks GA synthesis Reid and Potts, 1986
Is nana blocks GA synthesis Reid and Potts, 1986
na nana blocks ent-7a-hydr- Potts and Reid, 1983;
oxykaurenoic acid to Ingram and Reid
GA12-aldehyde (unpub.)
SnHr large produces polar GA Proebsting et al., 1978
response to
photo-
period
Arabidopsis thaliana ga-l non-germi- blocks GA synthesis Koornneef and
nating van der Veen, 1980
dwarf
4 James B. Reid

Action
ga-2 non-germi- blocks GA synthesis Koornneef and

nating van der Veen, 1980
dwarf
ga-3 non-germi- blocks GA synthesis Koornneef and
nating van derVeen, 1980
dwarf
ga-4 dwarf blocks GA intercon- Koornneef and
versions van der Veen, 1980
ga-5 dwarf blocks GA synthesis Koornneef and van
der Veen, 1980
Hordeum vulgare sin slender GA synthesis or uti1i- Foster, 1977
sation
br2 semi-dwarf blocks GA intercon- Suge,1983
versions
uz semi-dwarf blocks GA synthesis Suge,1972
GA-Iess dwarf lethal blocks GA synthesis Favret et al.. 1975
GA-ins semi-dwarf alters GA receptor Favret et al., 1975
gigas slender GA overproducer Favret etal., 1975
aleurone GA insensitive and Ho et al., 1980
mutants supersensitive
Lathyrus odoratus eJ dwarf blocks GA synthesis Magara, 1963
Oryza sativa dx dwarf blocks GA synthesis Murakami, 1972
prior to ent-kaurene
dy dwarf blocks GA20 to GAl Murakami, 1972
dwarfs GA-insensitive Harada and Vergarra,
1971
Triticum aestivum Rhtl semi-dwarf GA insensitive for Radley, 1970; Gale
stem elongation and Marshall, 1973
Rht2 semi-dwarf GA insensitive for Radley, 1970; Gale
stem elongation and Marshall, 1973
Rht3 dwarf reduces GA receptor Ho et aI., 1981 ;
concentration or Stoddart, 1984
affinity in both stem
and aleurone
Pharbitis nil dwarf blocks GA synthesis Ogawa, 1962;
Barendse and Lang,
1972
Citrullus lanatus dw-2 dwarf blocks GA synthesis Loy and Liu, 1974
Lycopersicon escu- ga non-germi- blocks synthesis prior Koornneef et al .,
len tum (Ve-335) nating to kaurene 1981;Zeevaart,1984
dwarf

Action
Ve-182 semi-dwarf partially blocks GA Zeevaart, 1984

synthesis
Ve-270 dwarf blocks GA-synthesis Zeevaart, 1984
between kaurenoic
acid and GAI2
sl2 stamenless blocks GA synthesis Sawhney, 1974
yg6 long-stem, GA overproducer Perez et al ., 1974
reduced
chlorophyll
Phaseolus vulgaris dwarf-I dwarf blocks GA synthesis Moh and Alan, 1967;
Proano and Greene,
1968
dwarf blocks GA synthesis Goto and Esashi, 1973
Prunus persica dwarf GA insensitive Wylie and Ryugo,
1971
Silene armeria s dwarf GA insensitive Wellensiek,1976
ef late flow- GA insensitive Wellensiek, 1976
ering
Lolium perenne d dwarf blocks GA synthesis Cooper, 1958
Ipomoea batatas dwarf GA insensitive Suge, 1979
Trifolium pratense dW1 dwarf GA sensitive Smith, 1974
Fragaria vesca arb long- removes block to GA Guttridge, 1973
stemmed synthesis
Ricinus communis dr slow germi- GA sensitive Shifriss, 1973
nating,
dwarf,
enhanced
maleness
Selection based upon changes in morphological characters can provide

only circumstantial evidence that a mutation is involved directly with GA
metabolism or sensitivity. For example, a GA metabolism mutant may be
indicated if an accurate phenocopy of the wild type is produced after
treatment of the mutant type with an active GA such as GA3 or GA I - The
hypothesis is further supported if treatment of the wild type with a GA syn-
thesis inhibitor such as AMO 1618 or PP 333 results in a phenocopy of the
mutant type. However, even where good application data are available,
direct proof that a GA metabolism mutant exists is difficult to obtain
because of the extremely low levels of GAs in vegetative tissue. Early
workers were confined to bioassays to quantify GA levels. While bioassays
have provided some of the classic results regarding the role of GAs (e. g.
Phinney, 1961) and are still the most appropriate starting point, they have
6 James B. Reid
at times been misleading and their use has been generally criticised
(Graebe and Ropers, 1978). This is partly due to their lack of specificity
and repeatability, although equally at fault has been the poor separation
techniques employed prior to the bioassay (Crozier and Durley, 1983) and
the poor choice of tissue extracted (i. e. the tissues extracted had not previ-
ously been shown to be a site in which the mutation was expressed).
Similar problems in comparing endogenous GA levels can occur with the
more definitive techniques such as GC-MS unless losses during purifi-
cation are accounted for and the correct tissue is used.
Proof that a GA metabolism mutant exists requires not only the identifi-
cation and quantification of the endogenous GAs but a demonstration of
differences in metabolism between the mutant and wild-type. For this
purpose GC-MS techniques and a supply of suitably labelled compounds
for use in metabolism studies are required. Even these techniques are
extended to their limits by the low levels of endogenous GAs in vegetative
tissues of wild types and are not sufficiently sensitive to examine the
"leakiness" of certain mutant types (e. g. Ingram et al., 1984). With these
types it may be necessary to turn to the more sensitive techniques of radio-
or enzyme-immunoassays. As with bioassays, problems may arise when
mass spectrometry (MS) or immunoassay techniques are employed. For
example, a comparison of tall (Le) and dwarf (Ie) peas using MS was
reported to show substantially more GAl in dwarf plants than in tall plants
(Keller and Coulter, 1982). The levels described were also far higher than
other reports (e. g. Ingram et al., 1984). The cause of these misleading
results is not completely clear but poor separation and purification were
partly at fault (see Gaskin et al., 1985). Even results from immunoassay
techniques may be at variance with results obtained from GC-MS tech-
niques (Atzorn and Weiler, 1983; Gilmour and MacMillan, 1984) sug-
gesting that further refinement of the specificity is required.
GA metabolism mutants may be subdivided into GA synthesis mutants
and GA breakdown or utilization mutants. Altered GA levels are indicative
of both groups. However, to date no mutants falling into the breakdown or
utilization categories have been proven but some of the proposed GA syn-
thesis mutants listed in Table 1 may well fall into these categories when
further work is done.
Designation as a GA insensitive mutant has up to now rested upon an
altered response to applied GA. They are a large and heterogeneous group
of mutants since they may influence any of the steps between reception of
the GA signal and the manifestation of the GA response. This group may
therefore include mutants which are not directly involved with the GAs.
For example, they include any mutant which makes the GAs non-limiting
for the response(s) being examined. These, at least potentially, include
factors such as insufficient substrates, limiting levels of other plant hor-
mones and limited uptake or transport systems within the plant. Many of
the GA insensitive mutants listed in Table 1 have not been examined in suf-
ficient detail to determine whether the GA insensitivity is of a direct,
primary nature or is an indirect effect via one of these mechanisms.
HOzC CH20H
'\
WCH20PP
GGPP \
_ _ _ _, ~CH20PP
'H CPP
-+-~
d,.
~H
ent-kaurene
MVA
Sterols Carotenolds 1
~
~~zOH ent-kaurenol
1
~+~-,
~;HO ~H
~
W" .
COzH
~
~~o ent-kaurenal
"'C~2'tt C02H
ent-kaurenoic
GA -aldehyde ent-70(-hydroxy-
'2 kaurenoic acid acid
d2~ or - \ da OH
~~
~:'H
COlH
~o
COzH
GA'2
GA 20 GA 2 .(inactive)
GA-insensitive mutants t,d,.d'

- ~ OH
Elongation
GA-receptor ~. H.
complex? ----HO .', .'
H~ CO,H
Other limiting factors GA, (active) GAB (inactive)
Fig. 1. The proposed sites of action of GA mutants in maize (d], db d 3 and d5), peas
(Ie, na) and rice (dy) in the early 13-hydroxylation gibberellin biosynthetic pathway
from mevalonic acid (MYA) to the biologically active product GAt. and the pos-
sible site(s) of action of GA insensitive mutants.
8 James B. Reid
III. G A Synthesis
An important impetus in determining the action of GA synthesis mutants

has been the determination of GA biosynthetic pathways in certain higher
plants. These have been reviewed recently by Sponsel (1983 a) and Mac-
Millan (1984). The pathway from mevalonic acid (MVA) to GA l2 -aldehyde
appears common to all the plants examined (see Hedden, 1983) (Fig. 1)
while after GA I2 -aldehyde several pathways differing largely in the pattern
and timing of hydroxylation occur (MacMillan, 1984). The one of most sig-
nificance to the discussion of GA mutants is the early 13-hydroxylation
pathway which has been demonstrated most extensively by Kamiya and
Graebe (1983) in cell free systems from pea seeds (Fig. 1). The point of
13-hydroxylation is not resolved but is likely to be between GA l2 and GA53
or GA I2 -aldehyde and GA5r aldehyde (shown as alternatives in Fig. 1). This
pathway probably occurs in the seeds of many higher plants since
13-hydroxylated GAs are native to a number of species (see Bearder, 1980).
It is also the probable pathway in other tissues including the shoots of peas
and maize (Davies et ai., 1982; Phinney and Spray, 1982; Gaskin et ai.,
1985) where the pathway after GA20 splits to produce either the inactive 2~
hydroxylated GA, GA29 , or the highly active 3~-hydroxylated GA, GAl
(Phinney and Spray, 1982; Ingram et ai., 1984; Spray et ai., 1984). Further
metabolism of GAl to GAg has been demonstrated (Davies and Rappaport,
1975). GA17 appears to be a side branch from GAl9 in this pathway (Davies
et ai., 1982; Kamiya and Graebe, 1983). A knowledge of the endogenous
GAs and the probable biosynthetic pathway is essential if application
studies are to provide evidence regarding the steps controlled by GA
mutants and the order in which they act. Early work in which the metab-
olism of, or response to, non-native GAs was examined did not provide
clear results (e. g. Crozier et ai., 1970; Musgrave and Kende, 1970).
IV. Internode Length
A. Synthesis Mutants
i) Maize
Thirty mutants influencing plant height have been reported in maize
(Phinney and Spray, 1982). The five dwarf mutants d], d2 , d3, dj and an] are
reported to block various steps in the GA biosynthetic pathway prior to the
formation of the active product, GAb since the application of appropriate
quantities of GA3 to any of these mutants results in a phenocopy of the
wild-type plant (Phinney, 1956) and all have reduced levels of GA-like sub-
stances when extracts from the shoots are bioassayed (Phinney, 1961).
Gene d j reduces the production of ent-kaur-16-ene (ent-kaurene) from
MVA to less than one quarter of that found in normals by reducing the B
activity of ent-kaurene synthetase (Fig. 1). This was demonstrated using a
cell-free system developed from young etiolated shoots of d5 and normal

maize seedlings (Hedden and Phinney, 1979). The major diterpene hydro-
carbon synthesised by the d5 system was ent-kaur-15-ene (ent-isokaurene)
regardless of whether [CI4]-MVA, [CI4]-geranylgeranylpyrophosphate
(GGPP) or [3H]-copalylpyrophosphate (CPP) was used as substrate. In
comparison, cell-free extracts from normal seedlings predominantly pro-
duced ent-kaurene (Hedden and Phinney, 1979). Ent-kaurene is a known
precursor to the GAs (Cross et al., 1964) but the isomer ent-isokaurene is
biologically inactive (Phinney and Spray, 1982) and not a precursor to the
GAs (Hedden et al., 1977). Both systems produce the other isomer as a
minor product. The reduction in ent-kaurene synthesis is almost certainly
the cause of the dwarf habit in d5 plants. The relatively early block in the
GA biosynthetic pathway by d5 is supported by the fact that the GA-pre-
cursors, ent-kaurene, ent-kaurenol, ent-kaurenoic acid as well as
GAlraldehyde, GAS3 and GA20 elicit a GA-like growth response when
applied to d5 seedlings (Katsumi et al., 1964; Phinney and Spray, 1982).
Gene d l blocks the 3~-hydroxylation of GA20 to GAl (Fig. 1) (Spray
et aI., 1984). Application of GA20 to d l plants results in a weak promotion of
elongation compared to other mutants (e. g. d2 , d3 or d5) and GA20 is only
one per cent as active as GAl (Phinney and Spray, 1982). When d l seedlings
were treated with [l3C, 3H]GA20 and the metabolites in the shoot identified
by GC-MS, only [l3C, 3H]GA29 and [13C,3H]GA29 -catabolite were found
(Spray et al., 1984). No evidence for the 3~-hydroxylated metabolites
[l3C,3H]GAI or [l3C,3H]GAg was found. In comparable feeds to normal (tall)
seedlings [l3C,3H]GA29 -catabolite and [l3C,3H]GAI were identified but no
[13C,3H]GAg or [l3C,3H]GA29 . Dilution of the [l3C] label by endogenous
[l2C]GAs was found in metabolites from both the normal and d l seedlings.
The cause of the absence of [l3C,3H]GA29 from the tall seedlings is unclear
but it was identified in similar feeds to d5 seedlings. No dilution of [l3C]
metabolites occurred in d5 seedlings confirming that the gene d5 blocks a
step prior to the formation of CwGAs. The late nature of the d l block is
supported by the fact that extracts from d l plants show several zones of
GA-like activity when bioassayed on d5 plants (Phinney and Spray, 1982).
Furthermore, in d5 - d l grafts the d5 shoots show a small but significant
increase in elongation compared with shoots in d5 - d5 control grafts
(Katsumi et al., 1983).
The steps in the GA-biosynthetic pathway blocked by genes d 2 , d 3 and
anI have not been determined with the same degree of certainty. Gene d 2
appears to block the carbon-7-oxidation step from GAsraldehyde to GAS3
and/or GA 12 -aldehyde to GA l2 (Fig. 1) since ent-kaurene, GA 12 -aldehyde
and GAsraldehyde show no activity when assayed on d2 plants while GAS3
and GA20 result in substantial elongation (Phinney and Spray, 1982;
Phinney, 1984). Similarly, gene d3 appears to block 13-hydroxylation from
GA l2-aldehyde to GAsraldehyde and/or GA l2 to GAS3 since d3 plants show
no response to kaurene or GAlraldehyde but respond well to
GAs3 -aldehyde, GAS3 and GA20 (Phinney and Spray, 1982; Phinney, 1984).
The site of action of anI has not been reported but it is possibly before
10 James B. Reid
kaurenol since Katsumi et al., (1964) reported that kaurenol could stim-
ulate leaf sheath elongation in an] plants.
ii) Rice
Of the wide range of GA sensitive mutants reported in rice, two recessive
dwarfs, dx and dy, have been examined in detail (Suge and Murakami,
1968; Murakami, 1972; Suge, 1978). Plants possessing gene dx (e. g. cv.
Tan-ginbozu) possess no GA-like substances in the shoots and show elon-
gation when treated with kaurene, kaurenol, kaurenoic acid or a wide
range of GAs (Murakami, 1972; Murofushi, 1983). This suggests that gene
dx blocks an early step (prior to kaurene) in the biosynthetic pathway
leading to the GAs (Fig. I) (Suge, 1978). Extracts from plants possessing
gene dy possess GA-like substances. The pattern found showed little dif-
ference from extracts from normal (Dy) plants (Suge and Murakami, 1968;
Murakami, 1972). Unlike dx plants, dy plants did not respond to kaurene,
kaurenol or kaurenoic acid and showed little response to non-3~-hydroxy
lated GAs such as GAl9 and GA20 (Murakami, 1972). Both dx and dy plants
respond well to 3~-hydroxylated GAs such as GAl and GA3 The biosyn-
thetic pathway for GAs has not been determined in rice although an early
13-hydroxylation pathway (Fig. 1) seems likely since GAl9 has been iden-
tified by GC-MS as the major GA in the shoots, roots and ears of rice
plants and GAl has been identified by GC-SICM (Kurogochi et al., 1979;
Suzuki et al., 1981 ; Murofushi, 1983). If this is the case, it seems likely from
application data that gene dy inhibits the 3~-hydroxylation of GA20 to GAl
(Fig. 1) (Murakami, 1972; Kurogochi et al., 1979). The genes dx and dyare
therefore closely comparable to the d 5 and d] genes in maize, respectively,
although detailed metabolic studies are required to confirm these sugges-
tions. Like the d5 and d] maize mutants the dx and dy dwarf rice mutants
have been used extensively as bioassays and the joint use of both mutants
has proved most useful in suggesting if GA-like fractions are 3~-hydroxyl
ated (Murakami, 1972; Potts et al., 1982).
iii) Barley
Many dwarf mutants occur in barley (Hordeum vulgare) although they
often differ in their cause (e. g. cell size, cell number, stem architecture,
etc., Blonstein, 1981). Suge (1983) discusses three non-allelic dwarf or semi-
dwarf mutants, UZ, br and br2. Gene br2 has been suggested to be equivalent
to the d] mutant of maize and the dy mutant in rice. However, the small dif-
ferences in the bioassay results and the crude separation techniques used
mean that firm evidence is still required. Gene uz has been reported to
reduce the level of GA-like substances in the shoot (Suge, 1972) but reports
by Kuraishi (1974) and Inouke et al., (1982) suggest that this gene possibly
has its primary action by inhibiting IAA synthesis, at least in dark-grown
coleoptiles. The levels of GA-like substances in br plants were similar to
those of normal varieties when bioassayed using the dwarf rice cv. Tan-
ginbozu. The results for these three mutants are at best inconclusive and
further detailed study is required.
Favret et al., (1975) and flopp et al., (1981) have examined two further
synthesis mutants. Their dwarf "GA-Iess" mutant becomes a phenocopy of
the normal type when treated with GA3 and contains no detectable GA-like
activity. The metabolism of GAl is not altered and it therefore appears to
be a synthesis mutant but its relationship to the mutants uz and br2 is not
clear. Of more interest is the gigas mutant (Favret et al., 1975). This giant
mutant is phenotypically similar to a normal plant treated with a large
quantity of an active GA. It appears to be a synthesis mutant since the GA
synthesis inhibitor CCC can reduce its growth and the "GA-less" mutant is
epistatic to the gigas mutant. While many GA-deficient mutants have been
described, few GA-overproducing mutants have been proposed and further
examination is warranted as it may allow the mechanisms controlling GA
synthesis to be determined.
Recent work with developing seeds of barley has shown they contain a
wide array of gibberellins (Gaskin et al., 1984) and a tentative scheme for
the metabolism of GAs in this species has been proposed (Gilmour et al.,
1984). This offers the hope that further progress on the action of the
mutants may be forthcoming.
iv) Peas
In the garden pea (Pisum sativum L.) over 40 mutants influencing plant
height have been reported (e. g. de Haan, 1927; Wellensiek, 1969; Sid-
orova, 1970; Blixt, 1972; Weber and Gottschalk, 1973). Eight non-allelic
mutants which have a marked influence on internode length have been
examined in some detail and four of these, Ie, na, Ih and Is, have been sug-
gested to influence gibberellin synthesis (Potts et al., 1982; Potts and Reid,
1983; Ingram et al., 1983, 1984; Reid and Potts, 1986).
Gene Ie has been subjected to many investigations (e. g. Brian and
Hemming, 1955; Lockhart, 1956; Jones and Lang, 1968; Kohler, 1970) and
results in a 70 per cent reduction in internode length compared to com-
parable Le (tall) plants (Blixt, 1972; Reid et al., 1983). Extracts from shoots
of tall (Le) plants contain the highly active GAb a GA not found in extracts
from the shoots of dwarf (Ie) peas (Potts et al., 1982; Davies et al., 1982;
Potts and Reid, 1983; Ingram et al., 1983). Application of GAl can mask
the difference between tall (Le) and dwarf (Ie) pea plants while only plants
possessing Le respond as well to GA 20 as to GAl (Ingram et al., 1983). This
suggests that the gene Ie is equivalent to the gene d J in maize and dy in rice
and probably acts by controlling the production of a 3~-hydroxylase which
allows the conversion of GA20 to GAl (Fig. 1). This was confirmed by
Ingram et al., (1984) who showed that [l3C,3H]GA20 is metabolised to
[13C,3H]GAb [l3C,3H]GAg and [l3C,3H]GA29 in the immature, shoot tissue of
Le plants. In contrast, [l3C,3H]GA29 and [13C,3H]GA29 -catabolite were the
only GAs produced in measurable quantities by plants homozygous for Ie.
However, Ie is probably a leaky mutant, as Ie plants show moderate elon-
gation if GA20 is applied (Ingram et al., 1983) and a trace of GAg was
detected in one dwarf (Ie) line by Ingram et al. (1984). A comparison of the
molecular ion currents of methyl TMS GAg and methyl TMS GA29 in the
12 James B. Reid
Ilfl. na (Ic
CON +GA, +GA,
Fig. 2. The response of the GA sensitive nana phenotype (Weibullsholm line 1766,
type line for gene na) and the GA-insensitive erectoides phenotype (John Innes line
1420, type line for gene lk) of peas to 10 J.1g of GAl applied to the third leaf after II
days growth. Line 1766 was produced by Professor S. J. Wellensiek and line 1420
was selected by Dr. P. Matthews.
dwarf gave a ratio of approximately 0.02 compared with a ratio of approxi-

mately 1.6 in the two tall (Le) lines examined (Ingram et al., 1984).
Gene na in peas has a more extreme effect on internode length than Ie
and results in the extremely short nana phenotype (Fig. 2) (Wellensiek,
1969; Reid et al., 1983). The young shoots from plants homozygous for na
do not possess detectable quantities of CwGAs since the p3C]-metabolites
from [13C,3H]GA2o feeds to na plants show no dilution with endogenous
p2C]GAs. In contrast, significant dilution was found in Na plants (Ingram
et al., 1984). The block caused by gene na is possibly prior to GA 19 and
G~ since in the rice seedling bioassay (cv. Tan-ginbozu) no GA-like sub-
stances were detected in extracts from the shoots of na plants (Potts and
Reid, 1983). G~ and GAI9 have been indicated by GC-MS from peaks in
the rice seedling bioassay from Na plants (Davies et al., 1982; Potts, unpub.
in Potts et al., 1985). Application data suggest that the block may be
between ent-7a-hydroxykaurenoic acid and GA 12 -aldehyde since na plants
showed no response to 1, 10 or 100 !!g of ent-kaurene, ent-kaurenol, ent-
kaurenal, ent-kaurenoic acid or ent-7a-hydroxykaurenoic acid but showed
a 90,280 and 380 per cent increase in internode length when treated with 1,
10 or 100 !!g of GA I2 -aldehyde, respectively (Ingram and Reid, unpub.).
However, further direct evidence is required as there is no evidence that
ent-kaurene, etc. were reaching a site within the plant where they could
Lh Ls
Fig. 3. The phenotype of two short-internode GA synthesis mutants of pea, Ih (in its
type line K 511) and Is (in its type line M26), compared with the parental cultivars
Torsdag (Lh) and Dippes gelbe Viktoria (Ls), respectively. K 511 and Torsdag were
provided by Dr. K. K. Sidorova and M 26 and Dippes gelbe Viktoria by Professor
W. Gottschalk.
elicit a biological response. In addition, GA 12 -aldehyde was not as effective

as GAl or GAzo in inducing elongation in na Le plants, the latter two com-
pounds inducing true phenocopies of isogenic Na plants.
Two further recessive dwarfing mutations in peas, Ih and Is (Fig. 3)
(Reid, 1986), appear to be GA synthesis mutants since phenocopies of
normal tall plants can be produced by appropriate treatment with the
native GA, GAl (Reid and Potts, 1986). Both Ih and Is plants contain little
or no GA-like substances when extracts from the shoots of 21 to 23 day old
plants were examined using the rice seedling bioassay (cv. Tan-ginbozu)
suggesting a block prior to G~ (Reid and Potts, 1986). The precise steps in
the GA biosynthetic pathway being affected will need to await the avail-
ability of suitably labelled intermediates.
Double recessive types, na Is, na Ih and Is Ih, possess shorter internodes
than the single recessive types (Reid, 1986). This suggests that the mutant
alleles are "leaky". However, it has not been possible to show the presence of
CwGAs in the shoots of na plants by bioassays (Potts and Reid, 1983) or by
dilution of [13C,3H]GA zo metabolites by endogenous [1zC]GAs using GC-MS
techniques (Ingram et al., 1984). Consequently, the genetic results suggest
that even though the levels of biologically active GAs are below the levels of
detection by these techniques they are still of biological significance.
14 James B. Reid
v) Other Species
Recent work by Zeevaart (1984) with three tomato mutants, Ve-182 (semi-
dwarf) and Ve-270 and Ve-335 (very short internodes) has produced similar
results to those found in maize, rice and peas. The semi-dwarf mutant
Ve-182 has qualitatively similar zones of GA-like activity to the parental cv.
"Money Maker" but the levels are much lower when determined by the d5
maize bioassay. Mutant Ve-182 is therefore a leaky mutant. The mutants
Ve-270 and Ve-335 possess no detectable GA-like activity. All mutants
become phenocopies of the wild-type if treated with biologically active
GAs. Feeds with GA precursors suggest that mutant Ve-335 may block GA
synthesis prior to kaurene while Ve-270 may block GA biosynthesis
between kaurenoic acid and GA I2 Perez et al., (1974) have suggested that
the long-stemmed tomato mutant yg6 is a GA overproducer and it may be
similar to the gigas mutant in barley.
In many other species dwarf mutants containing reduced levels of GAs
have been found (e. g. Lathyrus odoratus, Pharbitis nil, Phaseo/us vulgaris,
Arabidopsis thaliana, see Table I). Such mutants usually show a strong
growth response when treated with biologically active GAs. These data
provide circumstantial evidence that these mutants should be considered
GA synthesis mutants analogous to those described in detail for maize,
rice, peas and tomatoes. However, many of these results were collected
before the GA biosynthetic pathway was fully elucidated in any higher
plant and consequently the biochemical steps involved have not been
determined. Further work with these mutants is justified since it may help
to elucidate the GA biosynthetic pathway in these plants as well as provide
information regarding the control of elongation.
B. Insensitive Mutants
GA insensitive internode length mutants are common (see Table 1) and are
economically important since they are used extensively in agriculture (e. g.
wheat). The majority of insensitive mutants exhibit reduced internode
length [e. g. wheat (Allan et al., 1959), maize (Phinney, 1956)], but a few
exhibit increased internode length such as the slender phenotypes in peas
(de Haan, 1927) and barley (Foster, 1977). The two types will be considered
within the one group since the mode of action of both types of mutants
appears similar (e. g. Ho et al., 1981; Potts et al., 1985).
i) Wheat
Owing to their economic significance, the dwarfing genes in hexaploid
wheat (Triticum aestivum), Rht ], Rht 2 and Rht 3 have received the most
attention. Rht 3 displays either little or partial dominance over the
wild-type allele rht 3 (Morris et ai., 1972; Fick and Qualset, 1973) and is
more potent than the recessive genes Rht ] and Rht 2 (Fick and Qualset,
1973; McVittie et ai., 1978). Rht 3 and Rht ] are thought to be allelic and
located on chromosome 4A (Morris et ai., 1972; Gale and Marshall, 1976;
Gale and Law, 1977). Rht 2 is on chromosome 4D (Gale et al., 1975) and is
probably homoeologous with the Rht 3/Rht 1 locus (McVittie et al., 1978).
All three dwarfs are insensitive to GA in terms of elongation (Allan et al.,
1959; Gale and Marshall, 1973). At first this insensitivity was considered
an independent response (Hu and Konzak, 1974) controlled by the three
genes Gai 1, Gai 2 and Gai 3 closely linked to Rht 1, Rht 2 and Rht 3,
respectively (Gale and Law, 1977). However, it now seems likely that these
two responses may be pleiotropic effects of the same genes since no recom-
binants have been found even in large progenies (Gale et al., 1975; Fick
and Qualset, 1975; Gale and Law, 1977; McVittie et al., 1978). The GA
insensitivity of Rht 3 plants extends to all GA responses including a-
amylase production by the aleurone layer (Gale and Marshall, 1973; Fick
and Qualset, 1975; Ho et al., 1981). In contrast Rht 1 and Rht 2 plants do
not show GA insensitivity for a-amylase production (Radley, 1970; Gale
and Marshall, 1973; Fick and Qualset, 1975). The differences for Rht 1 and
Rht 3 are unexpected since alleles normally show similar sites of action
although differences may occur if the threshold for sensitivity differs
between tissues. The dominance relationships of Rht 1, Rht 3 and the
wild-type allele are also unexpected if Rht 1 and Rht 3 are allelic.
Radley (1970) examined the GA-like activity of Norin 10 type dwarf
wheat cultivars (probably possessing genes Rht 1 and Rht 2) using the
barley endosperm bioassay and found that the dwarfs contained more
GA-like activity than taIls regardless of whether germinating grains or
green seedlings were examined. Most of the activity was thought to be due
to GAl and it was suggested that there was a block to the utilisation of GA
in the dwarfs which resulted in its accumulation. Stoddart (1984) has
shown that the leaves of tall seedlings metabolise [3H]GA I rapidly to com-
pounds co-chromatographing on HPLC with [3H]GAg and a conjugate of
[3H]GAg. Rht 3 dwarf segregates from the same popUlation metabolised
[3H]GAI more slowly although the same products were produced. In both
cases 2~-hydroxylation of the active [3H]GAI to the inactive PH]GAg was
the initial metabolic modification. Radioimmunoassay indicated a 12 -15
times higher level of GA in the dwarf leaves, similar to the earlier findings
of Radley (1970). When the effects of disparate endogenous GA pool sizes
were taken into account the rates of 2~-hydroxylation were essentially com-
parable in tall and dwarf immature leaf samples, suggesting that insensi-
tivity in Rht 3 dwarfs is unrelated to enzymatic modification of the active
GAl' Instead, insensitivity probably relates to the ability to interact with
GAl (Stoddart, 1984). Ho et al. (1981) came to an essentially similar con-
clusion after comparing the physiology of the diverse GA responses in an
Rht 3 dwarf and a normal tall variety. In the dwarf, elongation of the
leaves, synthesis and release of hydrolytic enzymes and secretion of phos-
phate ions and reducing sugars by the aleurone layers were retarded. The
dwarf did not show a general slowdown in cellular metabolism, differ in
the uptake of GA3 or possess a different level of endogenous inhibitors
from the tall type. Differences between the cultivars in a-amylase pro-
duction were relatively greater at higher concentrations of GA3. This was
16 James B. Reid
used to suggest that in Rht 3 plants the level or activity of the GA receptor
for all responses was reduced (Ho et al., 1981).
ii) Other Species
The physiology of GA-insensitive dwarfs in maize, barley and peas (Fig. 2)
has not been examined in as much detail as in the dwarf wheats. However,
in barley and peas the levels of endogenous GA-like substances have been
examined and no qualitative changes have been found (Hopp et al., 1981;
Reid and Potts, 1986). Small quantitative differences do occur in peas but
these may be due to inherent differences in the ratios of leaf to stem tissue
extracted (Reid and Potts, 1986). All the available information is consistent
with the view that in the GA-insensitive dwarfs some step leading to the
GA response(s) at or after the GA-receptor site is limiting.
The GA insensitive slender phenotype in peas possesses long, thin
internodes, rapid germination, pale foliage, reduced branching, malformed
and abortive flowers, reduced seed set and the production of partheno-
carpic pods (de Haan, 1927; Reid et al., 1983). This set of pleiotropic char-
acteristics can be mimicked by the application of non-limiting quantities of
GA3 (Dalton and Murfet, 1975; Potts et al., 1985). The slender phenotype is
conferred by the combination of recessive genes la and cry' (see Reid et aI.,
1983) and is not altered by application of the GA synthesis inhibitors AMO
1618 or PP 333 (McComb and McComb, 1970; Potts et al., 1985). Further,
the slender gene combination fa cry' is epistatic to the GA synthesis
mutants fe (de Haan, 1927) and na (Potts et al., 1985). These results suggest
that the slender phenotype is not dependent on the level of endogenous
GAs (Brian, 1957; Potts et al., 1985). Direct support for this view was
obtained when the endogenous levels of GA-like substances were
examined using the rice seedling (cv. Tan-ginbozu) and lettuce hypocotyl
bioassays. Slender segregates (genotype fe fa cry' Na) were found to possess
qualitatively similar zones of GA-like substances to the dwarf segregates (Ie
fa Cry Na) although quantitatively the levels were lower in the slender
plants. Slender plants possessing the "GA-Iess" allele na possessed no
detectable GA-like substances even though they were phenotypically the
same as slender plants possessing Na (Potts et al., 1985). The slender gene
combination la cry' therefore allows the plant to act as if it is fully saturated
with GAs for growth regardless of the level of endogenous GAs, perhaps
by influencing a normally rate limiting step between the primary site of
perception of the GAs and the phenotypic response (Potts et al., 1985). The
exact mode of action is unknown but any hypothesis must take into
account the genetic evidence that La and Cry are near duplicate genes and
that the GA insensitive dwarfing gene lk is essentially epistatic to the gene
combination fa cry' (Reid, 1986). This may suggest la and cry' are operating
prior to the step controlled by lk and that la and cry' may act to release
some step which is normally repressed.
Very little is known about GA receptor sites or the steps between
reception and the production of a phenotypic response (see Stoddart and
Venis, 1980; Stoddart, 1983). This has been suggested as one of the major
problems facing plant hormone research (Vanderhoef and Kosuge, 1984).

The GA insensitive mutants may provide the necessary probes to examine
this problem especially where several such mutants occur in the same
species (e. g. maize, wheat and peas). Genetic evidence may allow the order
in which the mutants operate to be deduced.
v. Seed Dormancy
Recent work by Koornneef and co-workers has led to the isolation of non-
germinating dwarf mutants in Arabidopsis thaliana and tomato (Koornneef
and van der Veen, 1980; Koornneef et ai., 1981). In Arabidopsis 56 GA sen-
sitive dwarf mutants were isolated and shown to occur at five loci. Thirty
seven mutants at three of these loci, ga-l, ga-2 and ga-3, did not germinate
unless treated with GA (Koornneef and van der Veen, 1980). All GAs
tested, GA3 , G~+7' GA7 and GA9 , could induce germination with the
G~+7 mixture being the most effective. Without further GA treatment the
seedlings developed into dark green dwarfs but with weekly sprays of 10- 4
M G~+7 phenocopies of the wild-type were produced. Mutants with
similar phenotypes, except showing normal germination, also occurred at
loci ga-l, ga-2 and ga-3. Dwarf mutants at loci ga-4 and ga-5 were gen-
erally less extreme than mutants at ga-l, ga-2 and ga-3 (Koornneef and van
der Veen, 1980). Although the biochemical characterisation of these
mutants has not been reported it appears likely, based on the physiological
evidence, that they influence steps in GA synthesis. When the various
mutants at loci ga-l, ga-2 and ga-3 were further investigated a range from
an absolute to no GA requirement for germination was found. This sug-
gests that the alleles vary in their degree of "leakiness". The GA
requirement for germination is suggested to be much lower than for elon-
gation and flower development since in certain dwarfs germination can be
perfectly normal while length growth and flower development are substan-
tially altered (Koornneef and van der Veen, 1980). While this seems likely
the effectiveness of the genes in various tissues requires examination since
substantial differences between tissues can occur and certain GA-sensitive
dwarf mutants have been shown to be organ specific in their action (Potts
and Reid, 1983). The response of non-germinating tomato mutants is
similar (Koornneef et ai., 1981) and indicates the importance of GA for the
normal control of germination in these species. Along with ABA (see
Chapter 2 in this volume) and phytochrome mutants they should allow the
partial processes involved in this complex development process to be
examined. Auxotrophic mutants such as the non-germinating mutants are
rare in higher plants (Redei, 1975, and Chapter 10 in this volume) and
should be of considerable genetic use. Already they have been used to
allow the selection of germination revertants which appear to operate via
reduced ABA levels (Koornneef et ai., 1982, and Chapter 2 in this volume).
18 James B. Reid
VI. Flowering and Senescence
The involvement of GAs in the control of flowering has been the subject of
a recent review (Zeevaart, 1983). Reference here will therefore concentrate
on the garden pea where the genetic control of flowering and apical senes-
cence (or more correctly, cessation of apical growth) has been suggested to
operate via changes in GA metabolism (Proebsting et ai., 1978; Proebsting
and Heftmann, 1980). The genes at six established loci, If, e, sn, dne, hr and
veg interact to control the flowering behaviour of peas (see Murfet, 1985).
The complementary genes Sn and Dne confer the photoperiod response
(Barber, 1959; Murfet, 1971 a; King and Murfet, 1985). These two genes
have a wide range of pleiotropic effects so that the photoperiod response
can be recorded by observing anyone of a number of characters including
the node of first initiated flower, flowering time, apical senescence, number
of reproductive nodes, yield, the rate of flower development, vegetative
vigour and the production of lateral branches (Marx, 1968; Murfet, 1971 a,
1982; Reid and Murfet, 1984; Murfet and Reid, 1985). The most approp-
riate variable to use in examining the photoperiod response depends upon
the other flowering genes present since in certain combinations some of
these pleiotropic effects are masked (Murfet, 1971 a; Reid and Murfet,
1984). Grafting studies suggest that the gene combination Sn Dne regulates
the production of an inhibitor which delays flower initation (Murfet,
1971 b; Murfet and Reid, 1973; King and Murfet, 1985) and apical senes-
cence (Proebsting et ai., 1977). This inhibitor has a direct effect on apical
senescence (Reid, 1980) and is produced in both the leaves and cotyledons
(Murfet, 1971 a, 1985). The photoperiod response is mediated by phy-
tochrome (Reid and Murfet, 1977; Reid, 1979 a) and probably acts by regu-
lating some step in the Sn Dne pathway leading to the production of the
inhibitor. The presence of either sn and/or dne results in essentially day-
neutral plants (Murfet, 1971 a; King and Murfet, 1985). Gene Hrmagnifies
the photoperiod responses conferred by Sn Dne by prolonging the action of
the Sn Dne system in the leaves (Murfet, 1973; Reid, 1979 b). Consequently
plants possessing the combination Sn Dne Hr show pronounced photo-
period responses for at least certain characteristics.
Application studies have shown that GAs can influence the flowering
and apical senescence of peas (e. g. Barber etai., 1958; Wellensiek, 1969;
Davies et ai., 1977). The effects are generally small and do not suggest that
the flowering genes are operating directly by altering GA metabolism even
though the responses are strongly dependent on the genotype (Barber et ai.,
1958; Dalton and Murfet, 1975). However, two exceptions have been
reported. Davies et ai., (1977) reported that both GA3 and native GA20
could substantially delay the onset of apical senescence in the line G2
under long days (LD). Line G2 has genotype If E Sn Dne Hr Veg (Murfet,
1978) and thus possesses a large response to photoperiod for the number of
reproductive nodes and time and node of apical senescence (Marx, 1968).
Application of GA3 or GA20 under LD therefore has a similar delaying
effect to short days (SD) and led to the suggestion that the gene combi-
nation Sn Hr was responsible for GA production in the leaves in SD which

led to the delay in apical senescence (Davies et af., 1977; Proebsting et af.,
1978). Secondly, application of GA3 to line 24 plants (genotype Lf e Sn Dne
hr Veg) led to a small and possibly indirect delay in the node of first
initiated flower under continuous light but to a substantial and direct effect
under SD (Reid et af., 1977). To maximise this photoperiod response GA3
had to be present from an early stage of growth and therefore was
operating in an analogous fashion to the gene Hr (i. e. it only exerted a
large effect under conditions where the Sn Dne pathway would be
operating). Both pieces of evidence suggest a link between GAs and the
genes determining the size of the photoperiod response in peas.
In a more detailed study, Proebsting et af., (1978) reported that the
metabolism of [3H]G~ by leaves of G2 plants (Sn Dne Hr) was rapid in SD
and inhibited by LD. In SD a range of GAs more polar than G~ was pro-
duced. In contrast, cultivars homozygous for sn or hr (Marx's I types)
(Marx, 1968; Murfet, 1978) extensively metabolised [3H]G~ in both LD
and SD but the metabolites were reported to be different from those found
in G2 plants. In particular, one component, designated GAE, did not occur
in I types, being unique to G2 plants. Proebsting and Heftmann (1980)
found results similar to those obtained with G2 plants when feeds of
[3H]G~ were made to line G (Lf E Sn Dne Hr Veg) (Murfet, 1978) plants.
These plants show a large photoperiod response for the flowering node
(Marx, 1968), and can be induced to flower by one LD at the time of
treatment (eight weeks). When the levels of GA-like substances were
examined by the lettuce hypocotyl bioassay, extracts from the leaves of G2
plants grown under SD contained more GA-like activity in all three zones
of activity than similar plants exposed to 12 LD prior to harvest (Proeb-
sting et af., 1978). Transfer back to SD restored the levels of GA-like
activity. Extracts from the leaves of line 12 (If E Sn Dne hr Veg) grown under
SD contained a reduced level of GA-like activity compared to G2 plants in
the zone which co-chromatographed with GAE suggesting that the pre-
vention of apical senescence in G2 by SD may occur because the level of
GAE was sufficient to overcome any senescence stimulus associated with
fruit development (Proebsting et af., 1978). The level of GAE was suggested
to be controlled by genes Sn and Hr [Dne is present in these lines (King and
Murfet, 1985)]. In LD, or plants with sn or hr, GAE levels were too low to
prevent apical senescence (Proebsting et af., 1978) since the Sn Dne
pathway would be blocked at some point. GAE was originally suggested to
be GAl> but subsequent studies have not identified GAl from extracts of
line G2 or any other light grown dwarf (fe) line of peas (Davies et af., 1982;
Ingram, 1980; Sponsel, 1983 a). Identification by GC-MS of the
endogenous GAs from G2 plants led Ingram and Browning (1979) and
Davies et af., (1982) to suggest that the biological activity of GAE may be
due to GA I9 . However, GAl9 is not a likely metabolite of [3H]GA9. This
leaves the nature and particularly the significance of GAE in some doubt.
Ingram and Browning (1979) have shown that the levels of GAl9 and
GA20 are higher in developing seeds in SD than in LD relative to the level
20 James B. Reid
of GA29 . Ingram (1980) examined the levels of endogenous GAs by both

the lettuce hypocotyl bioassay and GCMS-MIM techniques in the devel-
oping seeds from nine pure lines with different flowering genotypes. GA19
and GA.,g levels were increased and GAl7 levels decreased in the two lines
carrying genes Sn and Hr but this response appeared independent of
photoperiod. In one line, the gene Hr alone appeared able to confer this
response. These results therefore fail to show a direct qualitative effect of
either Sn or Hr on GA levels in developing seeds. Further, in seedlings, no
effect of gene Sn on GA levels could be detected. Unfortunately isogenic
lines for genes Sn/sn and Hr/hr were unavailable during this work.
Potts (1982) and Potts, Reid and Murfet (unpub.) have examined the
levels of GA-like substances in approximately 20-day-old seedlings from
essentially isogenic lines for Sn/sn, Dne/dne and Hr/hr using the rice
seedling bioassay (cv. Tan-ginbozu). No qualitative differences in the levels
of GA-like activity were found and only minor quantitative differences
were apparent. In addition, photoperiod had only a minor quantitative
effect on the levels of GA-like activity in the Hr/hr and Dne/dne lines.
The results of Proebsting et al. (1978), Ingram (1980) and Potts (1982)
on the levels of endogenous GAs or GA-like activity do not provide strong
evidence for a direct control over a step in the GA biosynthetic pathway by
the genes Sn, Hr or Dne. The main differences recorded are possibly the
consequence of altered growth rates and developmental patterns caused by
these genes. The differences recorded in the levels of GA-like activity from
shoots possibly reflects the age of the tissue extracted (see Potts et al., 1982)
and changes associated with the onset of apical senescence. Potts et al.,
(unpub.) have shown that the level of GA-like activity drops in the shoot as
it approaches senescence.
This view is supported by evidence from the interaction of internode
length mutants, which are known to influence GA levels, and the flowering
genes. For example, the flowering phenotype of genotype Lf E Sn Dne Hr
Veg is the same regardless of whether Na or the "GA-deficient" mutant na
is present (Murfet, 1985) and genotype Lf E Sn Dne hr Veg shows normal
photoperiod responses regardless of whether Lh or Ih and Ls or Is are
present (Reid, unpub.). Further, segregation of the flowering genes Hr/hr,
Sn/sn and Dne/dne are not masked by the presence of the dwarfing gene Ie
(Barber, 1959; Murfet, 1971 a, 1973; King and Murfet, 1985). This is not to
suggest that the internode length genes do not have a pleiotropic effect on
flowering and vice versa (Barber, 1959; Dalton and Murfet, 1975) but
rather that even dramatic changes in the endogenous GA status of the plant
do not override the genetic control of flowering or the ability to respond to
photoperiod. Likewise, the GA synthesis inhibitor AMO 1618 cannot
override the effect of the flowering genes Sn and Hr although small
responses have been reported (Reid, 1976). If any of the flowering genes
Hr, Sn or Dne did operate by increasing the level of GAs it might be
expected that some increase in seedling internode length would occur.
N one of these genes leads to significant increases in internode length
(Barber, 1959; Potts, 1982; King, unpub.). These genetic and whole plant
physiological results all point to the photoperiod genes Sn and Dne and the
modifying gene Hr not operating directly via the control of steps in the GA
biosynthetic pathway. The biochemical mode of action of the flowering
genes in peas therefore remains a mystery. It is unfortunate that for early
studies of GA-Ievels and metabolism in the flowering genotypes isogenic
lines were unavailable and that the detection techniques were insufficient
to provide firm answers. In such situations less direct techniques such as
application studies and genetic recombinants are perhaps still capable of
providing the most valid answers.
Photoperiod control over GA metabolism has been firmly established
in other species (see Zeevaart, 1983) although mutants controlling these
responses have generally not been isolated. For example, spinach plants
exposed to LD possessed increased GA-like activity in the d5 maize
bioassay in one chromatographic zone and decreased activity in two other
zones compared with plants maintained in SD. The rate of turnover of GAs
appears to be increased under these LD conditions (Zeevaart, 1971). Iden-
tification of the native GAs by GC-MS suggested that an early 13-hydroxy-
lation biosynthetic pathway occurs in spinach (Metzger and Zeevaart,
1980a) (Fig. 1). Under SD the level of GA19 was high and that of GA20 was
low as measured by GC-SICM. After transfer to LD, the GA19 level
decreased and those of GA20 and GA29 increased, suggesting that photo-
period controls the conversion of GA 19 to GA20 (Metzger and Zeevaart,
1980 b). The change in GA metabolism occurred prior to the onset of stem
growth suggesting that GA20 levels may control stem elongation (Metzger
and Zeevaart, 1980b). Application of GA20 supports this suggestion (Zee-
vaart, 1983) as do feeds with [ZHJGAs3 since in SD only [ZHJG~4 and
[ZHJGA I9 were identified while after two LD only FHJGA20 was identified
(Gianfagna et al., 1983). However, the role that this step plays in the regu-
lation of flowering in this long-day plant (LDP) is still unclear.
In Silene armeria, another LDP where flower promotion is accom-
panied by stem elongation, mutants have been reported which affect the
time of flowering and the flowering response to GA3 as well as stem elon-
gation (Wellensiek, 1976). Cleland and Zeevaart (1970) found that AMO
1618 inhibited stem elongation but not flowering in this species suggesting
that flower formation is not under GA control. Transfer from SD to LD
was shown to increase the levels of GA-like substances in the d5 maize
bioassay. The studies of Silene mutants by Wellensiek (1976) suggest
responses not found using the single variety used by Cleland and Zeevaart
(1970). However, Suttle and Zeevaart (1979), working with the (GA insen-
sitive) stem elongation mutant, provide no evidence implicating the GAs
with flowering. If mutants influencing GA metabolism could be found in
spinach or Silene armeria the relationship between the photoperiod control
of GA metabolism, stem elongation and flowering might be resolved
finally in these LDP.
22 James B. Reid
VII. Site of Action of GA Mntants
The site of action of GA synthesis mutants has only been examined in a

preliminary manner. However, these results suggest that mutants may be of
considerable use in examining sites of GA production, transport and
action. For example, the pre-fruiting shoots of extremely short na pea
plants do not contain detectable levels of GAs (Potts and Reid, 1983;
Ingram et al., 1984) while the developing seeds contain similar levels of
GA-like substances to those found in seeds from normal Na plants (Potts
and Reid, 1983). Extracts from the pods at the time of contact of the deve-
loping seed also contain no GA-like activity corresponding to the major
zones of activity in the developing seeds (Potts et al., unpub.). These results
suggest little export of GA-like substances occurs from developing seeds, a
conclusion consistent with the observation that little increase in elongation
of the upper internodes occurs in na plants during fruiting. The high level
of GA-like substances in the developing seeds of na plants is also strong
evidence for the developing seeds as a site of GA-synthesis. This has been
shown previously in several species by detailed metabolic studies (e. g.
Kamiya and Graebe, 1983) but genetic evidence may provide a method for
analysing other potential sites of synthesis where the evidence from meta-
bolic studies is not as clear-cut.
Ingram et al. (1985) have examined the metabolism of [3HI3C]GA20 in
the roots following feeds to the shoots of na plants and found considerable
dilution of the [3HI3C]GAg-catabolite and the [3H13C]GA29-catabolite with
endogenous [l2C] compounds. In addition low levels of dilution were found
in [3H13C]GAg and [3H13C]GA29 in contrast to extracts from the shoots. This
suggests that na may be at least partially inoperative in the roots as well as
the developing seed. As a consequence no biologically active GAs would
be exported at detectable levels from the roots to the shoots contrary to the
hypothesis of Crozier and Reid (1971) that the roots convert biologically
inactive GAs from the shoots to active forms before reexport to the shoot.
Further, it suggests the roots are a site of GA synthesis. Major differences
in the range of metabolites present in the roots and shoots of na plants
were also observed. Perhaps the most significant difference was that the fl,
~-unsaturated ketone catabolites of GAg and GA29 were the predominant
metabolites in the roots but were relatively minor metabolites in the shoots
suggesting that the roots, through catabolism, may be important in the
control of the level of biologically-active GAs (Ingram et al., 1985). Sponsel
(1983b) has demonstrated a similar compartmentalisation of GA29 catab-
olism in the testa of maturing seeds. These results suggest that gene na may
not be an amorph. If this were so na might be expected to block GA syn-
thesis in all tissues. Many possibilities exist to explain these results
including alternative GA biosynthetic pathways, alternative enzymes (or
non-specific enzymes) in the tissues producing GAs or that na is a regu-
lator gene which is only operative in certain tissues.
Gene Le, which allows the 3~-hydroxylation of GA20 to GAl (Ingram et
al., 1984) also appears to be active in only specific sites within the pea plant
since GAl is found in the highest concentrations in the young expanding

apical portions of tall (Le) plants. It is found at much lower levels in
mature tissue (Potts et al., 1982; Potts and Reid, 1983) and has not been
identified in the developing seed (Eeuwens et aI., 1973; Gaskin et al., 1985).
This may partially account for the fact that the effect of the Le gene is not
graft transmissible (e. g. McComb and Mc Comb, 1970; Reid et al., 1983).
Even when Ie scions are grafted near actively growing Le apices no increase
in elongation occurs in the Ie scions suggesting that GAl may also be
immobilised or compartmentalised in some way within the apex (Reid,
unpub.). Jones and Lang (1968) have reported a difference in the diffusable
and extractable GA-like substances in the shoots of peas. From metabolism
studies using [3H]GA2o , the reduced level of GAl in mature tissues appears
to be at least partly due to a reduction in the rate of 3~-hydroxylation of
GA20 relative to 2~-hydroxylation of GA20 to GA29 (Ingram et al., 1985).
However, mature tissue can synthesise translocatable precursors of GAl
which may be of importance to shoot elongation in intact plants since Na
stocks can promote elongation in na scions (Reid et al., 1983). This genetic
evidence is consistent with evidence from other sources suggesting that the
young leaves are the main source of biologically active GAs within the
shoot (see Stoddart, 1983). The evidence from both na and Le plants clearly
shows precise and specific ontogenetic control over GA synthesis.
The lack of GAl in developing seeds of Le peas (Eeuwens et aI., 1973;
Gaskin et al., 1985) raises the question of the biological significance of the
high levels of other GAs found in developing seeds, especially when in
regard to stem elongation GAl has been suggested to be the only biologi-
cally active GA in peas, maize and rice (Phinney and Spray, 1982; Ingram
et al., 1984; Phinney, 1984; Murfet and Reid, 1985). In these plants, other
native GAs appear to possess activity only because of conversion to GAl.
AMO 1618 has been shown to dramatically reduce the level of GA-like
substances in developing seeds while having only a small effect on seed
development (Baldev et al., 1965). Neither Sponsel (1982) nor Ingram and
Browning (1979) could identify a clear role for the GAs in seed devel-
opment. Eeuwens and Schwabe (1975) and Sponsel (1982) suggest that
seed GAs are active in controlling pod development while Wareing and
Seth (1967) and Stoddart (1983) suggest that the high levels of GAs in devel-
oping seeds may be important in attracting nutrients. They do not appear
to be associated with the early growth of seedlings in peas since studies
with GA-synthesis inhibitors (Sponsel, 1983 b) and na plants (Reid, 1983)
show de novo synthesis is responsible for seedling elongation, at least from
node 1 onwards (counting from the cotyledons as zero). While it is possible
that different GAs are biologically active in the shoot and the developing
seed and pod this is perhaps unlikely and unnecessary to propose until
direct supportive evidence is available. The lack of a proven precise
function for the GAs in developing seeds and the absence at the current
levels of detection of GAl> even in Le seeds, might suggest that the GAs in
developing seeds are merely secondary metabolites without a key regu-
latory role. Clearly the physiological significance of GAs in developing
24 James B. Reid
seeds requires critical examination especially in view of the large number

of metabolic studies that have been based on this tissue.
The tissue specific nature of GA-synthesis mutants does not appear to
have been examined in other species. However, in maize comparisons can
be made between studies on seedlings and on the young tassels of maturing
plants. The levels of GAl and GAs were found to be very low in tassels from
normal plants, and conversion of [3H,13C]GA20 to [3H,13C]GAI and
[3H,BC]GAs could not be demonstrated in this tissue. [3H,13C]GA29 was the
only metabolite identified from such feeds in tassels from d 1, ds and normal
plants (Heupel et al., 1985). However, similar feeds to maize seedlings have
shown that GA20 is metabolised to GAl in the shoots of normal and ds seed-
lings but not in d 1 plants (Spray et al., 1984). This evidence suggests that,
like peas, the effect of GA synthesis mutants in maize may also be tissue
specific and illustrates the potential pitfalls that exist if the biochemical
action of a gene is examined in a tissue other than that in which it exerts its
primary effect.
The tissue specific nature of GA insensitive mutants is also well estab-
lished in wheat and barley. The Rht 3 gene causes insensitivity in all the
GA-mediated physiological processes examined (Gale and Marshall, 1973;
Ho et al., 1981) while the dwarfing genes from Norin 10 types (Rht 1 and
Rht 2) are reported to influence GA sensitivity only in the shoot and not in
the aleurone tissue (Gale and Marshall, 1973; Fick and Qualset, 1975). This
specificity of action may well prove to be commercially significant (Gale
and Hanson, 1982). In barley the reverse has been demonstrated where GA
insensitive aleurone mutants have been isolated in which the shoot appears
largely unaffected as no change in height was observed (Ho et al., 1980).
This suggests that at least certain steps are essential to all GA-mediated
responses while others are less basic and depend on the response and
tissue involved.
The influence of light and dark on dwarf mutants is one of the few
cases where GA mutants have been used to probe the GA responses to
environmental factors. In all cases examined [e. g. d 1, d3, ds in maize
(Sembdner and Schreiber, 1965), Ie, na, and la cry' in peas (Reid, 1983), and
Rht 3 and Rht 1 in wheat (Gale et al., 1975; Baroncelli et al., 1984)] the
effects of the genes are expressed in darkness as well as in light, regardless
of whether the mutants are GA synthesis or GA sensitivity types. This sug-
gests that GA levels or sensitivity are limiting for growth in both the light
and dark and that in none of the above cases does darkness allow the steps
blocked by these mutants to be completely overcome. The term 'physio-
logical dwarf was coined to describe the situation in which dwarfness
occurred in the light but not in the dark (see Jones, 1973) and was based on
results with tall (Le) and dwarf (Ie) cultivars of peas (Lockhart, 1956).
However the genetic evidence outlined above argues that the effects of
light and GAs are, at least partly, independent of each other (Reid, 1983;
Baroncelli et al., 1984). Gaskin et al. (1985) recently reported that the
shoots from a dark grown dwarf (Ie) cultivar of peas (cv. Progress No.9)
contain similar levels of GAl to those from a tall (Le) cultivar (cv. Alaska)
and show similar elongation. Unfortunately, these cultivars are unrelated

and differ in many other respects, including the levels of other GAs. The
ratio of GAl to GA20 or GA29 is greater in the tall cultivar than in the dwarf
cultivar, consistent with the known effect of the Le gene in light grown
plants. These results therefore do not resolve the question of physiological
dwarfism in peas but do highlight the need for the use of isogenic lines or
segregating progenies if the effect of a particular gene difference is to be
examined.
VIII. Conclusions
The biochemical site of action has now been determined for several GA
synthesis mutants influencing internode length (Fig. 1). These results
provide some of the strongest evidence available regarding the control of a
development process by GA levels and point towards GAl being the only
active GA in controlling elongation in maize, peas and rice (Phinney,
1984). These mutants provide precise tools for blocking GA synthesis at
known steps and they may be of importance in determining the sites of GA
synthesis, action and transport since preliminary information suggests that
they show precise ontogenetic and tissue specificity. GA synthesis mutants
influencing seed dormancy, onset of flowering, apical senescence and sex
determination have also been proposed (Table 1) but cover only a minority
of the developmental processes and environmental responses suggested to
be controlled or influenced by the GAs. Further examination of the pleio-
tropic effects of proven GA synthesis mutants and the isolation and
selection of further mutants may provide the best avenue available to
define the regulatory role of endogenous GAs. For example, we presently
have detailed information regarding the levels and biosynthetic pathways
in developing seeds but lack basic information regarding their biological
function.
GA insensitive mutants are common, particularly for internode length
(Table 1). Evidence points to at least some of these mutants controlling the
GA receptor site or limiting steps between this point and the manifestation
of a GA response rather than by influencing GA metabolism (Stoddart,
1984; Potts et ai., 1985). The mechanism of action of GAs has proven a dif-
ficult area of study (Stoddart, 1983) and has recently been highlighted as
an area of potential significance (Vanderhoef and Kosuge, 1984). The GA
insensitive mutants offer as yet unexploited probes for examining this
process especially where mutants are available to influence a sequence of
steps in the one species.
Developmental mutants in higher plants, and GA mutants in particular,
have not been exploited to the same degree as mutants in bacteria and
lower plants possibly owing to the more complex developmental pathways
involved and the partitioning of the problems between disciplines.
However, their potential as invaluable research tools in plant biology has
been demonstrated (e. g. Phinney, 1961) and provided a multidisciplinary
26 James B. Reid
approach, involving genetics, anatomy, physiology and biochemistry is

adopted, they offer one of the best methods available to gain a complete
understanding of facets of plant development.
Acknowledgements
I wish to thank Drs. T. J. Ingram, I. C. Murfet, W. C. Potts and J. J. Ross

for helpful discussions during the development of my views on this topic
and/ or helpful comments on the manuscript, Mrs. T. Grabek for technical
support and the Australian Research Grants Scheme for financial
assistance.
IX. References
Allan, R. E., Vogel, O. A, Craddock, J. C., 1959: Comparative response to gibber-

ellic acid of dwarf, semi-dwarf and tall winter wheat varieties. Agron. J. 51,
737-740.
Atzorn, R., Weiler, E. W., 1983: The role of endogenous gibberellins in the for-
mation of a-amylase by aleurone layers of germinating barley. Planta 159,
289-299.
Baldev, B., Lang, A, Agatep, A 0., 1965: Gibberellin production in pea seeds
developing in excised pods: effect of growth retardant AMO 1618. Science 147,
155-157.
Barber, H. N., 1959: Physiological genetics of Pisum. II. The genetics of photoperi-
odism and verna1isation. Heredity 13, 33-60.
Barber, H. N., Jackson, W. D., Murfet, I. C., Sprent, J. I., 1958: Gibberellic acid
and the physiological genetics of flowering in peas. Nature (London) 182, 1321.
Barendse, G. W. M., Lang, A, 1972: Comparison of endogenous gibberellins and
the fate of applied radioactive gibberellin AI in a normal and a dwarf strain of
Japanese morning glory. Plant Physiol. 49, 836-841.
Baroncelli, S., Lercari, B., Cionini, P. G., Cavallini, A, D'Amato, F., 1984: Effect of
light and gibberellic acid on coleoptile and first-foliage-leaf growth in durum
wheat (Triticum durum L. Desf.). Planta 160, 298-304.
Bearder, J. R., 1980: Plant hormones and other growth substances - their back-
ground, structure and occurrence. In: MacMillan, J. (ed.), Hormonal Regu-
lation of Development. I. Molecular Aspects of Plant Hormones (Encyclo-
paedia of Plant Physiology, New Series, Vol. 9), pp. 9-112. Berlin - Hei-
delberg - New York: Springer-Verlag.
Bearder, J. R., 1983: In vivo diterpenoid biosynthesis in Gibberella fujikuroi: the
pathway after ent-kaurene. In: Crozier, A (ed.), The Biochemistry and Phys-
iology of Gibberellins, Vol. 1., pp. 251-387. New York: Praeger.
Blixt, S., 1972: Mutation genetics in Pisum. Agri Hortique Genetica 30, 1-293.
Blonstein, A D., 1981: Developmental differences between dwarfing genes in
barley. In: Asher, M. J. C. (ed.), Barley Genetics IV (Proceedings of the Fourth
International Barley Genetics Symposium, Edinburgh 1980), pp. 566-570.
Edinburgh: Edinburgh Univ. Press.
Brian, P. W., 1957: The effects of some microbial metabolic products on plant
growth. Symp. Soc. Exp. BioI. 11, 166-181.
Brian, P. W., Hemming, H. G., 1955: The effect of gibberellic acid on shoot growth
of pea. Physiol. Plant. 8, 669-681.
Cleland, C. F., Zeevaart, J. A D., 1970: Gibberellins in relation to flowering and
stem elongation in the long-day plant Silene armeria. Plant Physiol. 46,
392-400.
Cooper, J. P., 1958: The effect of gibberellic acid on a genetic dwarf in Lolium
perenne. New Phytol. 57, 235-238.
Cross, B. E., Galt, R H. B., Hanson, J. R, 1964: The biosynthesis of the gibber-
ellins. Part I. (-)-Kaurene as a precursor of gibberellic acid. J. Chern. Soc.
295-300.
Crozier, A, 1983: The Biochemistry and Physiology of Gibberellins, Vols. 1 and 2,
p. 568 and p. 452. New York: Praeger.
Crozier, A., Durley, R. c., 1983: Modern methods of analysis of gibberellins. In:
Crozier, A (ed.), The Biochemistry and Physiology of Gibberellins, Vol. 1,
pp. 485-560. New York: Praeger.
Crozier, A, Kuo, C. c., Durley, R c., Pharis, R P., 1970: The biological activities
of 26 gibberellins in nine plant bioassays. Can. J. Bot. 48, 867 -877.
Crozier, A, Reid, D. M., 1971: Do roots synthesize gibberellins? Can. J. Bot. 49,
967-975.
Dalton, P. J., Murfet, I. C., 1975: The effect of gibberellic acid and genotype Ie la
cry on flowering in peas. Pisum Newsl. 7, 5-7.
Davies, L. J., Rappaport L., 1975: Metabolism of tritiated gibberellins in d-5 dwarf
maize. II. PH] gibberellin AI and PH] gibberellin A 3 , and related compounds.
Plant Physiol. 56, 60-66.
Davies, P. J., Emshwiller, E., Gianfagna, T. J., Proebsting, W. M., Noma, M.,
Pharis, R. P., 1982: The endogenous gibberellins of vegetative and reproductive
tissue of G2 peas. P1anta 154, 266-272.
Davies, P. J., Proebsting, W. M., Gianfagna, T. J., 1977: Hormonal relationships in
whole plant senescence. In: Pilet, P. E. (ed.), Plant Growth Regulation
(9th International Conference on Plant Growth Substances, 1976, Lausanne),
pp. 273-280. Berlin - Heidelberg - New York: Springer-Verlag.
de Haan, H., 1927: Length factors in Pisum. Genetica 9, 481-497.
Eeuwens, C. J., Gaskin, P., Macmillan, J., 1973: Gibberellin A 20 in seed of Pisum
sativum L. cv. Alaska. Planta 115, 73- 76.
Eeuwens, C. J., Schwabe, W. W., 1975: Seed and pod wall development in Pisum
sativum L. in relation to extracted and applied hormones. J. Exp. Bot. 26, 1-14.
Favret, E. A., Favret, G. C., Malvarez, E. M., 1975: Genetic regulatory mechanisms
for seedling growth in barley. In: Barley Genetics III (Proc. 3rd Int. Barley
Genet. Symp.), pp. 37 -42. Garching, W. Germany.
Fick, G. N., Qualset, C. 0., 1973: Genes for dwarfness in wheat, Triticum aestivum
L. Genetics 75, 531-539.
Fick, G. N., Qualset, C. 0., 1975: Genetic control of endosperm amylase activity
and gibberellic acid responses in standard-height and short-statured wheats.
Proc. Nat. Acad. Sci., U.S.A 72, 892-895.
Foster, C. A., 1977: Slender: an accelerated extension growth mutant of barley.
Barley Genet. Newsl. 7, 24-27.
Gale, M. D., Hanson, P. R, 1982: The plant breeding potential of genetic variation
in cereal phytohormone systems. In: McLaren, J. S. (ed.), Chemical Manipu-
lation of Crop Growth and Development (Proceedings of Easter School of
Agricultural Science, Univ. of Nottingham), pp. 425-449. London: Butter-
worths.
28 James B. Reid
Gale, M. D., Law, C. N., 1977: The identification and exploitation of Norin 10
semi-dwarfing genes. Plant Breeding Inst., Cambridge. Annu. Report. pp.
21-35.
Gale, M. D., Law, C. N., Marshall, G. A., Worland, A J., 1975: The genetic control
of gibberellic acid insensitivity and coleoptile length in a 'dwarf wheat. Her-
edity 34, 393-399.
Gale, M. D., Marshall, G. A, 1973: Insensitivity to gibberellin in dwarf wheats.
Ann. Bot. 37, 729-735.
Gale, M. D., Marshall, G. A., 1976: The chromosomal location of Gai 1 and Rht 1,
genes for gibberellin insensitivity and semi-dwarfism, in a derivative of Norin
10 wheat. Heredity 37, 283-289.
Gaskin, P., Gilmour, S. J., Lenton, J. R., MacMillan, J., Sponsel, V. M., 1984:
Endogenous gibberellins and kaurenoids identified from developing and germi-
nating barley grain. J. Plant Growth Regul. 2, 229-242.
Gaskin, P., Gilmour, S. J., MacMillan, J., Sponsel, V. M., 1985: Gibberellins in
immature seeds and dark-grown shoots of Pisum sativum. Gibberellins iden-
tified in the tall cultivar Alaska in comparison with those in the dwarf Progress
No.9. Planta 163, 283-289.
Gianfagna, T., Zeevaart, J. A D., Lusk, W. J., 1983: Effect of photoperiod on the
metabolism of deuterium-labelled gibberellin AS3 in spinach. Plant Physiol. 72,
86-89.
Gilmour, S. J., Gaskin, P., Sponsel, V. M., MacMillan, J., 1984: Metabolism of gib-
berellins in immature barley grain. Planta 161, 186-192.
Gilmour, S. J., MacMillan, J., 1984: Effect of inhibitors of gibberellin biosynthesis
on the induction of a-amylase in embryoless caryopses of Hordeum vulgare cv.
Himalaya. Planta 162, 89-90.
Goto, N., Esashi, Y., 1973: Diffusible and extractable gibberellins in bean coty-
ledons in relation to dwarfism. Physiol. Plant. 28, 480-489.
Graebe, J. E., Ropers, H. J., 1978: Gibberellins. In: Letham, D. S., Goodwin, P. B.,
Higgins, T. J. V. (eds.), Phytohormones and Related Compounds - A Compre-
hensive Treatise. Vol. I. The Biochemistry of Phytohormones and Related Com-
pounds, pp. 107 -204. Amsterdam: Elsevier/North-Holland.
Guttridge, C. G., 1973: Stem elongation and runnering in the mutant strawberry,
Fragaria vesca L. arborea Staudt. Euphytica 22, 357 -361.
Harada, J., Vergarra, B. S., 1971: Response of different rice varieties to gibberellins.
Crop Sci. 11, 373-374.
Hedden, P., 1983: In vitro metabolism of gibberellins. In: Crozier, A (eds.), The
Biochemistry and Physiology of Gibberellins, Vol. 1, pp. 99-149. New York:
Praeger.
Hedden, P., MacMillan, J., Phinney, B. 0., 1978: The metabolism of the gibber-
ellins. Annu. Rev. Plant Physiol. 29, 149-192.
Hedden, P., Phinney, B. 0., 1979: Comparison of ent-kaurene and ent-isokaurene
synthesis in cell-free systems from etiolated shoots of normal and dwarf-5
maize seedlings. Phytochemistry 18, 1475-1479.
Hedden, P., Phinney, B. 0., MacMillan, J., Sponsel, V. M., 1977: Metabolism of
kaurenoids by Gibberella fujikuroi in the presence of the plant growth retardant,
N, N, N-trimethyl-l-methyl-(2',6',6'-trimethy1cyc1ohex-2'-en-l'-yl)prop-2-enyl-
ammonium iodide. Phytochemistry 16, 1913 -1917.
Heupel, R. C., Phinney, B. 0., Spray, C. R., Gaskin, P., MacMillan, J., Hedden, P.,
Graebe, J. E., 1985: Native gibberellins and metabolism of [l4C] gibberellin AS3
and of [l7-13C, 17 -3H 2] gibberellin A20 in tassels of Zea mays. Phytochemistry 24,
47-53.
Ho, T. D., Nolan, R. C., Shute, D. E., 1981: Characterisation of a gibberellin-insen-
sitive dwarf wheat, D6899. Plant Physiol. 67, 1026-1031.
Ho, T. D., Shih, S., Kleinhofs, A., 1980: Screening for barley mutants with altered
hormone sensitivity in their aleurone layers. Plant Physiol. 66, 153 -157.
Hoad, G. V., 1983: Gibberellin bioassays and structure-activity relationships. In:
Crozier, A. (ed.), The Biochemistry and Physiology of Gibberellins, Vol. 2,
pp. 57-94. New York: Praeger.
Hopp, H. E., Favret, G. C., Favret, E. A., 1981: Control of barley development
using dwarf mutants. In: Induced Mutants as a Tool for Crop Improvement, p.
243. Vienna: IAEA.
Hu, M. L., Konzak, C. F., 1974: Genetic association of gibberellic acid insensitivity
and semi-dwarfing in hexaploid wheat. Annu. Wheat Newsl. 20, 184-185.
Ingram, T. J., 1980: Gibberellins and reproductive development in peas. Ph. D.
Thesis, University of East Anglia, U. K.
Ingram, T. J., Browning, G., 1979: Influence of photoperiod on seed development
in the genetic line of peas G2 and its relation to changes in endogenous gibber-
ellins measured by combined gas chromatography - mass spectrometry. Planta
146,423-432.
Ingram, T. J., Reid, J. B., MacMillan, J., 1985: Internode length in Pisum sativum. L.
The kinetics of growth and PH] gibberellin A20 metabolism in genotype na Le.
Planta 164, 429-438.
Ingram, T. J., Reid, J. B., Murfet, I. C., Gaskin, P., Willis, C. L., MacMillan, J.,
1984: Internode length in Pisum. The Le gene controls the 3~-hydroxylation of
gibberellin A 20 to gibberellin A j Planta 160,455-463.
Ingram, T. J., Reid, J. B., Potts, W. C., Murfet, I. C., 1983: Internode length in
Pisum. IV. The effect of the Le gene on gibberellin metabolism. Physiol. Plant.
59,607-616.
Inouke, M., Sakurai, N., Kuraishi, S., 1982: Growth regulation of dark-grown
dwarf barley coleoptile by the endogenous IAA content. Plant Cell Physiol. 23,
689-698.
Jones, R. L., 1973: Gibberellins: their physiological role. Annu. Rev. Plant Physiol.
24,571-598.
Jones, R. L., Lang, A., 1968: Extractable and diffusible gibberellins from light- and
dark-grown pea seedlings. Plant Physiol. 43, 629-634.
Kamiya, Y., Graebe, J. E., 1983: The biosynthesis of all major pea gibberellins in a
cell-free system from Pisum sativum. Phytochemistry 22, 682-689.
Katsumi, M., Foard, D. E., Phinney, B. 0., 1983: Evidence for the translocation of
gibberellin A3 and gibberellin-like substances in grafts between normal, dwarf j
and dwarf 5 seedlings of Zea mays L. Plant Cell Physiol. 24, 379-388.
Katsumi, M., Phinney, B. 0., Jefferies, P. R., Hendrick, C. A., 1964: Growth
response of the d-5 and an-J mutants of maize to some kaurene derivatives.
Science 144, 849-850.
Keller, P. L., Coulter, M. W., 1982: The relationship of endogenous gibberellins to
light-regulated stem elongation rates in dwarf and normal cultivars of Pisum
sativum L. Plant Cell Physiol. 23, 409-416.
King, W., Murfet, I. c., 1985: Flowering in Pisum: A sixth locus, Dne. Ann. Bot., 56,
835-846.
Kohler, D., 1970: The effect of red light on the growth and gibberellin-content of
pea seedlings. Z. Pflanzenphysiol. 62, 426-435.
30 James B. Reid
Koornneef, M., Jorna, M. L., Brinkhorst - van der Swan, D. L. C., Karssen, C. M.,
1982: The isolation of abscisic acid (ABA) deficient mutants by selection of
induced revertants in non-germinating gibberellin sensitive lines of Arabidopsis
thaliana (L.). Heynh. Theor. Appl. Genet. 61, 385-393.
Koornneef, M., van der Veen, J. H., 1980: Induction and analysis of gibberellin sen-
sitive mutants in Arabidopsis thaliana (L.). Heynh. Theor. Appl. Genet. 58,
257-263.
Koornneef, M., van der Veen, J. H., Spruit, C. J. P., Karssen, C. M., 1981: Isolation
and use of mutants with an altered germination behaviour in Arabidopsis
thaliana and tomato. In: Induced Mutants as a Tool for Crop Plant
Improvement 251, 227-232. Vienna: IAEA-SM.
Kuraishi, S., 1974: Biogenesis of auxin in the coleoptile of a semi-brachytic barley,
uzu. Plant Cell Physiol. 15, 295-306.
Kurogochi, S., Murofushi, N., Ota, Y., Takahashi, N., 1979: Identification of gib-
berellins in the rice plant and quantitative changes of gibberellin A 19 throughout
its life cycle. Planta 146, 185-191.
Lockhart, J. A, 1956: Reversal of light inhibition of pea stem growth by gibber-
ellins. Proc. Nat. Acad. Sci., U.S.A 42, 841-848.
Loy, J. B., Liu, P. B. W., 1974: Response of seedlings of a dwarf and a normal strain
of watermelon to gibberellins. Plant Physiol. 53, 325-330.
McComb, A. J., McComb, J. A., 1970: Growth substances and the relation between
phenotype and genotype in Pisum sativum. Planta 91, 235-245.
MacMillan, J., 1980: Hormonal Regulation of Development. I. Molecular Aspects
of Plant Hormones (Encyclopedia of Plant Physiology, New Series, Vol. 9).
Berlin - Heidelberg - New York: Springer-Verlag.
MacMillan, J., 1984: Analysis of plant hormones and metabolism of gibberellins.
In: Crozier, A, Hillman, J. R. (eds.), The Biosynthesis and Metabolism of Plant
Hormones (Society of Experimental Biology Seminar, Series 23), pp. 1-16.
Cambridge: Cambridge University Press.
McVittie, J. A, Gale, M. D., Marshall, G. A, Westcott, B., 1978: The intra-chromo-
somal mapping of the Norin 10 and Tom Thumb dwarfing genes. Heredity 40,
67-70.
Magara, J., 1963: Notes on the possible role of endogenous "gibberellins" in the
determining of monofactorial dwarfism in dwarf sweet peas (Lathyrus odoratus
L.) and in mutants d 1 and d s of maize. Ann. Physiol. Veg. 5, 249-261.
Marx, G. A., 1968: Influence of genotype and environment on senescence in peas,
Pisum sativum L. BioScience 18, 505-506.
Metzger, J. D., Zeevaart, J. A D., 1980a: Identification of six endogenous gibber-
e1lins in spinach shoots. Plant Physiol. 65, 623-626.
Metzger, J. D., Zeevaart, J. A. D., 1980b: Effect of photoperiod on the levels of
endogenous gibberellin in spinach as measured by combined gas chromatog-
raphy-selected ion current monitoring. Plant Physiol. 66, 844-846.
Moh, C. C., Alan, J. J., 1967: The response of a radiation-induced dwarf bean
mutant to gibberellic acid. Turrialba 17, 176-178.
Morris, R., Schmidt, J. W., Johnson, V. A, 1972: Chromosomal location of a
dwarfing gene in "Tom Thumb" wheat derivative by monosomic analysis. Crop
Sci. 12,247-249.
Murakami, Y., 1972: Dwarfing genes in rice and their relation to gibberellin biosyn-
thesis. In: Carr, D. J. (ed.), Plant Growth Substances, 1970, pp. 166-174.
Berlin - Heidelberg - New York: Springer-Verlag.
Murfet, I. C., 1971 a: Flowering in Pisum. Three distinct phenotypic classes deter-
mined by the interaction of a dominant early and a dominant late gene. Her-
edity 26, 243-257.
Murfet, I. C., 1971 b: Flowering in Pisum: reciprocal grafts between known geno-
types. Aust. J. BioI. Sci. 24, 1089-1101.
Murfet, I. C., 1973: Flowering in Pisum. Hr, a gene for high response to photo-
period. Heredity 31, 157 -164.
Murfet, I. C., 1978: The flowering genes Lf, E, Sn and Hr in Pisum: their rela-
tionship with other genes, and their descriptions and type lines. Pisum Newsl.
10,48-52.
Murfet, I. C., 1982: Flowering in the garden pea: expression of gene Sn in the field
and use of multiple characters to detect segregation. Crop Sci. 22, 923-926.
Murfet, I. c., 1985: Pisum sativum L. In: Ha1evy, A. H. (ed.), Handbook of Flow-
ering, Vol. IV, pp. 97-126. Boca Raton, Florida: CRC Press.
Murfet, I. C., Reid, J. B., 1973: Flowering in Pisum: evidence that gene Sn controls
a graft-transmissible inhibitor. Aust. J. BioI. Sci. 26, 675-677.
Murfet, I. c., Reid, J. B., 1985: The control of flowering and internode length in
Pisum. In: Hebblethwaite, P. D., Heath, M. c., Dawkins, T. C. K. (eds.), The
Pea Crop: a Basis for Improvement, pp 67-80. London: Butterworths.
Murofushi, N., 1983: Life cycle regulation in rice by endogenous plant hormones.
In: Miyamoto, J. (ed.), Pesticide Chemistry. Human Welfare and the Envi-
ronment (Proceedings of the 5th International Congress of Pesticide Chemistry,
1982, Kyoto), pp. 21-28. Oxford: Pergamon Press.
Musgrave, A., Kende, H., 1970: Radioactive gibberellin As and its metabolism in
dwarf peas. Plant Physiol. 45, 53-55.
Ogawa, Y., 1962: Quantitative differences of gibberellin-like substances in normal
and dwarf varieties of Pharbitis nil. Chois. Bot. Mag. Tokyo 75, 449-450.
Perez, A. T., Marsh, H. V., Lachman, W. H., 1974: Physiology of the yellow-green 6
gene in tomato. Plant Physiol. 53, 192-197.
Phinney, B. 0., 1956: Growth response of single-gene dwarf mutants in maize to
gibberellic acid. Proc. Nat. Acad. Sci., U.S.A. 42, 185-189.
Phinney, B. 0., 1961: Dwarfing genes in Zea mays and their relation to the gibber-
ellins. In: Klein, R. M. (ed.), Plant Growth Regulation, pp. 489-501. Ames,
Iowa: Iowa State College Press.
Phinney, B. 0., 1984: Gibberellin AI. dwarfism and the control of shoot elongation
in higher plants. In: Crozier, A., Hillman, J. R. (eds.), The Biosynthesis and
Metabolism of Plant Hormones (Society of Experimental Biology Seminar,
Series 23), pp. 17-41. Cambridge: Cambridge University Press.
Phinney, B. 0., Spray, C., 1982: Chemical genetics and the gibberellin pathway
in Zea mays L. In: Wareing, P. F. (ed.), Plant Growth Substances 1982,
pp. 10 1-11 O. London Academic Press.
Potts, W. C., 1982: The involvement of gibberellins with the internode length and
flowering genotypes of Pisum. Ph. D. Thesis, University of Tasmania, Australia.
Potts, W. C., Reid, J. B., 1983: Internode length in Pisum. III. The effect and inter-
action of the Na/na and Le//e gene differences on endogenous gibberellin-like
substances. Physiol. Plant. 57, 448-454.
Potts, W. C., Reid, J. B., Murfet, I. C., 1982: Internode length in Pisum. I. The effect
of the Le//e gene difference on endogenous gibberellin-like substances. Physiol.
Plant. 55, 323-328.
Potts, W. C., Reid, J. B., Murfet, I. C., 1985: Internode length in Pisum. Gibberellins
and the slender phenotype. Physiol. Plant. 63, 357 -364.
32 James B. Reid
Proano, V. A., Greene, G. L., 1968: Endogenous gibberellins of a radiation induced

single gene dwarf mutant of bean. Plant Physiol. 43, 613-618.
Proebsting, W. M., Davies, P. J., Marx, G. A., 1977: Evidence for a graft-transmis-
sible substance which delays apical senescence in Pisum sativum L. Planta 135,
93-94.
Proebsting, W. M., Davies, P. J., Marx, G. A., 1978: Photoperiod-induced changes
in gibberellin metabolism in relation to apical growth and senescence in genetic
lines of peas (Pisum sativum L.). Planta 141, 231-238.
Proebsting, W. M., Heftman, E., 1980: The relationship of PH]GA9 metabolism to
photoperiod induced flowering in Pisum sativum L. Z. Pflanzenphysiol. 98,
305-309.
Radley, M., 1970: Comparison of endogenous gibberellins and response to applied
gibberellin of some dwarf and tall wheat cultivars. Planta 92, 292-300.
Redei, G. P., 1975: Induction of auxotrophic mutations in plants. In: Ledoux, L.
(ed.), Genetic Manipulations with Plant Material, pp. 329-349. New
York - London: Plenum Press.
Reid, J. B., 1976: Regulation of Flowering in Pisum. Ph. D. Thesis, University of
Tasmania, Australia.
Reid, J. B., 1979a: Red-far-red reversibility of flower development and apical sen-
escence in Pisum. Z. Pflanzenphysiol. 93, 297 -301.
Reid, J. B., 1979b: Flowering in Pisum: the effect of age on the gene Sn and the site
of action of gene Hr. Ann. Bot. 44, 163-173.
Reid, J. B., 1980: Apical senescence in Pisum: a direct or indirect role for the flow-
ering genes. Ann. Bot. 45, 195-201.
Reid, J. B., 1983: Internode length genes in Pisum. Do the internode length genes
affect growth in dark-grown plants? Plant Physiol. 72, 759-763.
Reid, J. B., 1986: Internode length in Pisum. Three further loci, Ih, Is and Ik. Ann.
Bot., 57, in press.
Reid, J. B., Dalton, P. J., Murfet, I. C., 1977: Flowering in Pisum: does gibberellic
acid directly influence the flowering process? Aust. J. Plant Physiol. 4,
479-483.
Reid, J. B., Murfet, I. C., 1977: Flowering in Pisum: the effect of light quality on
genotype If e Sn Hr. J. Exp. Bot. 28, 1357 -1364.
Reid, J. B., Murfet, I. C., 1984: Flowering in Pisum: a fifth locus, Veg. Ann. Bot. 53,
369-382.
Reid, J. B., Murfet, I. C., Potts, W. c., 1983: Internode length in Pisum. II. Addi-
tional information on the relationship and action of loci Le, La, Cry, Na and
Lm. J. Exp. Bot. 34, 349-364.
Reid, J. B., Potts, W. c., 1986: Internode length in Pisum. Two further mutants, Ih
and Is, with reduced gibberellin synthesis, and a gibberellin insensitive mutant,
Ik. Physiol. Plant., in press.
Sawhney, V. K., 1974: Morphogenesis of the stamenless-2 mutant in tomato. III.
Relative levels of gibberellins in the normal and mutant plants. J. Exp. Bot. 25,
1004-1009.
Sembdner, G., Schreiber, K., 1965: Activities of gibberellin A3 , As and A6 on the
maize-dwarf mutants d], d 3 and d s in the dark. Flora, oder Allg. Bot. Z. Abt. A
156, 359-363.
Shifriss, 0.,1973: The drooping syndrome of Ricinus. J. Hered. 64, 351-355.
Sidorova, K. K., 1970: The study of allelism in phenotypically identical mutants of
pea in connection with the law of homologous series in hereditary variability.
Genetika 6, 23-35.
Smith, R. R., 1974: Inheritance of a gibberellin-responsive dwarf mutant in red

clover. Euphytica 23, 597 -600.
Sponsel, V. M., 1982: Effects of applied gibberellins and napthylacetic acid on pod
development in fruits of Pisum sativum L. cv. Progress No.9. 1. Plant Growth
Regul. 1, 147-152.
Sponsel, V. M., 1983 a: In vivo gibberellin metabolism in higher plants. In: Crozier,
A (ed.), The Biochemistry and Physiology of Gibberellins, Vol. 1, pp. 151-250.
New York: Praeger.
Sponsel, V. M., 1983 b: The localisation, metabolism and biological activity of
gibberellins in maturing and germinating seeds of Pisum sativum cv. Progress
No.9. Planta 159, 454-468.
Spray, C., Phinney, B. 0., Gaskin, P., Gilmour, S. 1., MacMillan, 1., 1984: Internode
length in Zea mays L. The dwarf-l mutant controls the 3~-hydroxylation of gib-
berellin A 20 to gibberellin A). Planta 160,464-468.
Stoddart, 1. L., 1983: Sites of gibberellin biosynthesis and action. In: Crozier, A.
(ed.), The Biochemistry and Physiology of Gibberellins, Vol. 2, pp. 1-55. New
York: Praeger.
Stoddart, 1. L., 1984: Growth and gibberellin-A) metabolism in normal and gibber-
ellin-insensitive (Rht 3) wheat (Triticum aestivum L.) seedlings. Planta 161,
432-438.
Stoddart, 1. L., Venis, M. A, 1980: Molecular and subcellular aspects of hormone
action. In: MacMillan, 1. (ed.), Hormonal Regulation of Development. I.
Molecular Aspects of Plant Hormones (Encyclopaedia of Plant Physiology,
New Series, Vol. 9), pp. 445-510. Berlin - Heidelberg - New York: Springer-
Verlag.
Suge, H., 1972: Effect of uzu (uz) gene on the level of endogenous gibberellins in
barley. lpn. 1. Genet 47, 423-430.
Suge, H., 1978: The genetic control of gibberellin production in rice. lpn. 1. Genet.
53, 199-207.
Suge, H., 1979: Gibberellin relationships in a dwarf mutant of sweet potato. lpn. 1.
Genet. 54, 35-42.
Suge, H., 1983: Gibberellin relationships in a dwarf mutant of barley: brachytic
(br2). lpn. 1. Genet. 58, 555-566.
Suge, H. Murakami, Y., 1968: Occurrence of a rice mutant deficient in gibberel-
lin-like substances. Plant Cell Physiol. 9, 411-414.
Suttle, 1. C., Zeevaart, 1. A D., 1979: Stem growth, flower formation and
endogenous gibberellins in a normal and a dwarf strain of Silene armeria.
Planta 145, 175-180.
Suzuki, Y., Kurogochi, S., Murofushi, N., Ota, Y., Takahashi, N., 1981: Seasonal
changes of GAl. GA)9 and abscisic acid in three rice cultivars. Plant Cell
Physiol. 22, 1085-1093.
Trewavas, A., 1981: How do plant growth substances work? Plant Cell Environ. 4,
203-228.
Vanderhoef, L. N., Kosuge, T., 1984: The molecular biology of plant hormone
action: research directions of the future (Workshop summaries - II), p. 40.
Rockville: Am. Soc. Plant Physiol.
Wareing, P. F., Seth, A. K., 1967: Aging and senescence in the whole plant. Symp.
Soc. Exp. BioI. 21, 543-558.
Weber, V. E., Gottschalk, W., 1973: Die Beziehungen zwischen ZellgroBe und
InternodienHinge bei strahleninduzierten Pisum-Mutanten. Beitr. BioI. Pflanz.
49, 101-126.
34 James B. Reid
Wellensiek, S. J., 1969: The physiological effects of flower forming genes in peas.
Z. Pflanzenphysiol. 60, 1388-1402.
Wellensiek, S. J., 1976: A genetical look at flower formation in Silene armeria L. In:
Jacques, R. (ed.), Etudes de Biologie Vegetale, Hommage au Professor Pierre
Chouard, pp. 301-312. Paris: Louis-Jean.
Wylie, A., Ryugo, K., 1971: Diffusible and extractable growth regulators in normal
and dwarf shoot apices of peach, Prunus persica Botsch. Plant Physiol. 48,
91-93.
Zeevaart, J. A. D., 1971: Effects of photoperiod on growth rate and endogenous
gibberellins in the long-day rosette plant spinach. Plant Physiol. 47, 821- 827.
Zeevaart, J. A. D., 1983: Gibberellins and flowering. In: Crozier, A. (ed.), The Bio-
chemistry and Physiology of Gibberellins, Vol. 2, pp. 333-374. New York:
Praeger.
Zeevaart, J. A. D., 1984: Gibberellins in single gene dwarf mutants of tomato. Plant
Physiol. Suppl. 75, 186.
Chapter 2
Genetic Aspects of Abscisic Acid
Maarten Koomneef
Department of Genetics, Agricultural University, 53, Generaal Foulkesweg,

NL-6703 BM Wageningen, The Netherlands
With 3 Figures
Contents
I. Introduction
A. Abscisic Acid as a Plant Hormone
B. Biosynthesis and Metabolism of ABA
C. Genetic Aspects of ABA
II. ABA-Deficient Mutants
A. Isolation, Genetic and General Phenotypic Characteristics
B. Pleiotropic Effects of ABA-Deficient Mutants
C. Biochemistry of ABA-Deficient Mutants
D. The Use of ABA Mutants in Water Relations Research
E. The Use of ABA Mutants in Seed Physiology
F. ABA Deficiency in Relation to Other Physiological Effects
III. Mutants Affecting ABA Sensitivity
IV. Genetic Differences in ABA Accumulation
V. Conclusions
VI. References
I. Introduction
A. Abscisic Acid as a Plant Hormone

Abscisic acid (ABA) is a naturally occurring plant hormone, probably
present in all higher plants. Its discovery in the sixties, and its chemical
structure have been described in several reviews (Addicott and Carns,
1983; Milborrow, 1984). ABA may be involved, often as an inhibitor, in
many physiological processes such as abscission, bud- and seed dormancy,
elongation growth, stomatal opening, root growth, geotropism, fruit ri-
pening and senescence (Walton, 1980; Addicott and Carns, 1983; Mil-
36 M. Koornneef
borrow, 1984). In many cases, however, the role of ABA could not be estab-
lished conclusively, mainly because of the inadequate experimental
approaches. In addition, experiments examining the correlation between
endogenous ABA levels and physiological effects, or those involving the
exogenous application of ABA etc. never provide more than circumstantial
evidence, and compartmentation as well as tissue- and time-specific differ-
ences in hormone sensitivity complicate the interpretation of results.
Methods of manipulating endogenous hormone levels would provide more
direct approaches (Karssen, 1982) but specific chemical inhibitors, such as
there are for gibberellins, are not known for ABA. The use of isogenic
genotypes differing in endogenous ABA content provides plant physiolo-
gists with an important tool to elucidate the regulatory function of this
compound. This genetic approach has been successfully applied to study
the relation of ABA with stomatal closure and seed dormancy. Mutants
also contributed to research on the pathway by which ABA is synthesized
in the plant.
B. Biosynthesis and Metabolism ofABA

ABA is a sesquiterpenoid derived from mevalonic acid (MVA). Two
pathways have been suggested (Milborrow, 1983):
1. The direct pathway, where ABA is synthesized from a C-15 precursor,
presumably farnesyl pyrophosphate.
2. The indirect pathway, where ABA is formed via cleavage of a C-40 pre-
cursor such as violaxanthin, a xanthophyll that is derived from caro-
tenes. Xanthoxin is probably the ABA precursor derived from violax-
anthin (Fig. 1).
Experiments by Creelman and Zeevaart (1984) showed that stress-
induced ABA synthesized in an 18 0 2 atmosphere contained a labelled
oxygen atom only in the carboxyl group, providing strong evidence that at
least stress-induced ABA is derived from a C-40 precursor. The arguments
in favour of the direct pathway come from experiments where 14C phytoene
and 3H-MVA were fed to avocado fruits; only 3H was found in ABA (Mil-
borrow, 1983). Creelman and Zeevaart (1984) have critically discussed
these experiments.
ABA can be rapidly metabolized to phaseic acid (PA), for example
upon rehydration of wilted leaves (Zeevaart, 1980) and PA thereafter is
reduced to dihydrophaseic acid (DPA) (Fig. 1). ABA and its direct metabo-
lites can be conjugated with glucose (Walton, 1980; Milborrow, 1983,
1984). Although in some systems PA has been reported to have hormone
activity (e. g. Uknes and Ho, 1984), in general only ABA is found to be
active in biological test systems (Walton, 1980).
C. Genetic Aspects ofABA

Genetic aspects of ABA have been reviewed by Quarrie (1983), who
emphasizes the potential use of ABA in agriculture. The present review
Abscisic Acid 37
Mevalonic acid(MVA)
!
l
Carotenes
OH
HO
Violaxanthin
~
HO~ c!o Xanthoxin
Absc is i c ac i d (ABA)
C02H Phaseic acid (PA)
C02 H Di hyd rophaseic acid (DPA)
Fig. 1. The proposed "indirect" biosynthetic pathway for ABA and the major
pathway for its catabolism
38 M. Koornneef
concentrates on the use of genetic variation for ABA in basic plant

research. The genetic variation that is available derives from mutation
induction experiments but may be found also among natural populations
and domesticated cultivars of a specific species. In general, more extreme
variation will be found among mutants than among "natural variants" as
for the latter more severe limitations exist for the changed gene on survival
and yield.
Table 1. Predicted Phenotypes of ABA Mutants
Effects compared to wild type

Effect of on: Sensitivity
Process Results in
mutation ABA-induced ABA level to ABA
responses
Biosynthesis decrease deficiency 0

Biosynthesis increase overproduction + + ?
Catabolism decrease overproduction + + ?
Catabolism increase deficiency
Reception decrease hyposensitivity 0
Reception increase hypersensitivity + 0 +
Uptake, decrease hyposensitivity 0 0
transport
Uptake, increase hypersensitivity + 0 +
transport
o : comparable to wild type

+ : increased compared to wild type
- : reduced compared to wild type
A crucial step in genetic studies is the ability to recognize genetic vari-

ation. Table 1 summarises the predicted phenotypes of possible ABA
mutants. However, this scheme is compromised by the rapid changes in
ABA levels due to stress. Thus, enhanced ABA levels may be a secondary
effect of, for example, a mutation that reduces the root system and,
therefore, increases stress and ABA levels. Furthermore, because ABA acts
on several different physiological processes, extensive biochemical analysis
may be required to establish the exact lesion. Phenotypic changes merely
provide preliminary indications of possible mutations.
II. ABA-Deficient Mutants
A. Isolation, Genetic and General Phenotypic Characteristics

ABA-deficient mutants thus far have been recognized because they showed
excessive wilting and/or reduced seed dormancy. These effects were
shown to be reversed by exogenous ABA and correlated with greatly
reduced endogenous ABA levels. In mutants of Arabidopsis (Koomneef et
Abscisic Acid 39
al., 1982), potato (Quarrie, 1982) and tomato (Tal and Nevo, 1973;
Koomneef et al., 1982, 1985) both water relations and seed dormancy are
affected. Effects of the wilty mutation in pea isolated by L. Cruger of the
Del Monte Corporation, San Leandro, U. S. A. (see Marx, 1976) on seed
dormancy have not been reported (Donkin et al., 1983; Wang et al., 1984).
In the viviparous mutants of maize, which are seedling 1ethals, wilting is
not obvious, although J. D. Smith (pers. comm.) mentions that they are
hypersensitive to water stress.
All ABA-deficient mutants so far described are monogenic reces-
sives. Tomato mutants with mutations at three different loci, sitiens
(sit), flacca (fic) and notabilis (not) were isolated by Stubbe (1957, 1958,
1959) in the progeny of X-ray treated seeds. Another mutant at the sit
locus induced by ethylmethanesulphonate was isolated by Van der
Veen and Bosma (Koornneef et al., 1985). Droopy (dr) mutants of
potato (Solanum tuberosum L group phureja) were found in a cross of
two clones (CPC 2862 x 2863) that were apparently heterozygous for dr
(Simmonds, 1965).
All maize mutants with reduced ABA levels are carotenoid deficient
and map to at least seven different loci viz.: viviparous 2, 5 and 9 (vp2, vp5,
vp9), pink scutellum (ps= vp7), albescent (al= y3), white seedling 3 (w3) and
yellow (y9). The mutants arose from different sources and many alleles have
been independently isolated (Robertson, 1955, 1975). These seedling
letha1s are recognized on segregating ears (thus on heterozygous wild type
mother plants) as white seeds, many of which germinate precociously on
the ear. Mutant vp8, which is a viviparous green dwarf, may be an ABA-
deficient mutant (Smith et al., 1978), but no ABA determinations have been
published.
In Arabidopsis thaliana (Koornneef and van der Veen, 1980) and in
tomato (Koornneef, et al., 1985), gibberellin-deficient mutants do not ger-
minate without exogenous gibberellin (see also Chapter 1 in this volume).
By selecting for germinating seeds amongst the progeny of mutagen-treated
non-germinating (ga-l) Arabidopsis seeds, mutants were isolated that were
reverted at least for the non-germination trait. Genetic analysis showed this
reversion to be due to a mutation in a suppressor gene segregating inde-
pendently from the ga-l gene.
Table 2 shows the scheme whereby this suppressor mutant was isolated.
The fact that in tomato the double mutant sitw/sitw, ga-l/ga-l does not ger-
minate without gibberellin indicates that the procedure used to isolate
ABA mutants in Arabidopsis is not applicable to every species. However,
the ability of ABA mutants (e. g. sit in tomato) to germinate in conditions
that are adverse to the germination of the wild type, such as in a germi-
nation medium with high osmotic potential, or under continuous far-red
irradiation, provides an alternative scheme for the direct selection of such
non-dormant mutants (Koornneef et al., 1985).
Mutants that have not yet been characterized with respect to ABA, but
which have phenotypes similar to some of the ABA mutants mentioned
above, include a non-dormant, carotenoid-free, white mutant of Helianthus
40 M. Koornneef
Table 2. Isolation of Isogenic, Recessive GA and ABA Mutants of Arabidopsis

thaliana
STAGE 1
Generation Phenotype Genotype

Mo Wild type Ga-l/Ga-J, Aba/Aba
~ EMSa ~ ~
MJ Wild type Ga-l/ga-l, Aba/Aba
~ selfing ~ ~
M2 GAMutant ga-l/ga-l, Aba/Aba
Germination: GA dependent
Plants: dwarfs
STAGE 2
Mo GAmutant ga-l/ga-l, Aba/Aba
~ EMsa ~ ~
MJ GAmutant ga-l/ga-l, Aba/aba
~ selfing ~ ~
M2 Revertant ga-l/ga-l, aba/aba
Germination:GA independent
Plants: dwarfs
STAGE 3
Revertant ga-l/ga-l, aba/aba
x x
x Wild type Ga-l/Ga-J, Aba/Aba
~ ~
Wild type Ga-l/ga-l, Aba/aba
~ ~
ABA Mutant Ga-l/Ga-l, aba/aba
Germination: GA independent
Plants: wilting, normal height
aEMS = ethylmethanesulphonate; 10 mM applied for 24 hr in darkness at 24 cC.
annuus L. (Wallace and Habermann, 1958) and viviparous mutants in

Avena nuda L. (Cummings et al., 1978).
It should be emphasized that not all wilting mutants and mutants with
reduced seed dormancy are ABA deficient. For example, the wilted (wi)
mutant of maize (Postlethwait and Nelson, 1957) and wilting dwarf (wd) in
tomato (Alldridge, 1964) wilt because anomalous vessel elements restrict
water movement through the plants. Excessive stomatal opening, which is
characteristic for all wilty ABA mutants, can result from the disturbed
regulation of ion uptake mechanisms (Tal et al., 1976) as found, for
example, in the scabrous diminutive mutant of Capsicum annuum (Tal et
aI., 1974).
Abscisic Acid 41
A reduction in seed dormancy may also be caused by factors other than

ABA deficiency, as shown by the germination behaviour of Arabidopsis
seed coat mutants (Koornneef, 1981). The correlation between reduced
ABA content and low sprouting resistance (reduced dormancy) in barley
found by Goldbach and Michael (1976) is difficult to accept as a causal
relationship as only two non-isogenic cultivars were compared and joint
segregation of ABA levels and sprouting resistance in segregating prog-
enies was not studied.
B. Pleiotropic Effects ofABA-Deficient Mutants

True pleiotropism is inferred from the joint occurrence of the same phe-
notypic characters in independently isolated mutations at the same locus
and from the absolute association of the complex phenotype in segregating
populations. Both the presence and the absence of particular pleiotropic
effects in ABA mutants can provide useful information about the function
of ABA. However, some pleiotropic effects are very indirect and have no
causal relation to the hormone. For example, semi-dwarfness in ABA-defi-
cient mutants is due to their disturbed water relations and should not be
interpreted as a direct effect of ABA on cell division or cell elongation.
Tomato ABA mutants show leaf epinasty, swelling of the upper stem and
rooting along the stem and it is difficult to explain these characteristics as
direct ABA effects. Tal et al. (1979) suggest that these traits are caused by
auxin overproduction which subsequently enhances ethylene production.
ABA deficiency is clearly the primary effect because all characters are
reversed by exogenous ABA. In the dr mutant of potato (Quarrie, 1982)
and aba mutants of Arabidopsis, such auxin-like effects have not been
observed. An additional pleiotropic effect of aba mutants in Arabidopsis is
the reduction of mucilage excretion (Karssen et al., 1983).
All viviparous mutants of maize with reduced ABA levels are charac-
terized by pale yellow endosperms in genotypes that otherwise would be
yellow, and by white or almost white seedlings (Robertson, 1975).
Defective carotenoid biosynthesis is the primary lesion in these mutants
leading to the multiple defects described (Fong et ai., 1983). That this is a
direct causal relationship is shown by the application of fIuridone (a pyr-
idone inhibitor of carotenoid biosynthesis) to the wild type, which results
in phenocopies of the vp5 mutants (Smith et al., 1983).
C. Biochemistry ofABA-Deficient Mutants

The wilty mutants of tomato, potato, pea and Arabidopsis still contain low
levels of ABA. For the three tomato mutants, ABA levels in non-stressed
leaves were found to be 12-15 % of wild type for sit, 17 -26 % for fic and
31-49 % for not. These percentages refer respectively to data published by
Tal and Nevo (1973) and by Neill and Horgan (1985) and correlate approx-
imately to the severity of the symptoms in the different mutants. Such a
correlation has also been observed in the different alleles of the aba locus
42 M. Koornneef
of Arabidopsis. The most viable mutant, aba3 (G4), contains 10-26 % of the
wild-type value in immature and mature seeds (Koornneef et ai., 1982),
whereas in aba l (A26) ABA was below the detection level (Koornneef et ai.,
1982, 1984). Still less vigorous aba alleles (e. g. abct=A73) are known
which are almost lethal due to their extreme wilting and withering
(Koornneef et ai., 1982). ABA levels in pea and potato are about 11 % of
that of wild type in stressed leaves (Quarrie, 1982; Wang et ai., 1984).
In potato (Quarrie, 1982), sit and fie tomato (Neill and Horgan, 1985),
pea (Wang et ai., 1984) and aba3 Arabidopsis (Zeevaart, pers. comm.) ABA
does not accumulate in response to water stress as it does in the wild type.
It may be that the wilty mutants are already stressed under normal condi-
tions and have reached their maximal ABA level. Alternatively, stress-
induced ABA cannot be produced in these mutants, suggesting different
pathways in turgid and water-stressed leaves. However, because the muta-
tions in tomato (Groot and Karssen, pers. comm.) and Arabidopsis
(Koornneef et ai., 1982; Karssen et ai., 1983) affect ABA levels in seeds and
leaves in a similar way, a leaf-specific pathway seems unlikely.
All carotenoid-deficient mutants of maize have reduced but significant
levels of ABA in embryo tissue, the values being variable and age dependent
(1. D. Smith, pers. comm.). Published data (Brenner et ai., 1977; Smith et ai.,
1978; Robichaud et ai., 1980) range from 7 - 70 % of that of the wild type.
However, ABA is not detectable in w3, vp5 and vp7 seedlings and roots
(Moore and Smith, 1985). In view of these data and preliminary results of in
vitro cultures of kernels (1. D. Smith, pers. comm.), the ABA found in mutant
embryos produced on heterozygous wild-type mother plants is very probably
of maternal origin. All deficient maize mutants accumulate specific car-
otenoid precursors but not zeaxanthin (Robertson, 1975; Fong et al., 1983).
The blocks in carotenogenesis suffered by these mutants are presented in Fig.
2 after Fong et ai., 1983). The relationships between the blocks in the car-
otenoid pathway and the effects on ABA levels, both in mutants (Moore and
Smith, 1985) and after application of the carotenoid inhibitor fluridone
(Moore and Smith, 1984), provide strong arguments for the C-40 pathway of
ABA synthesis. As carotenoid deficiency indirectly affects many aspects of
plant metabolism, for example leading to excessive photo-oxidation of chi or-
ophylls, it may be argued that ABA deficiency is an indirect effect of car-
otenoid deficiency. ABA synthesis may need a functional chlorophyll system
as suggested by Quarrie and Lister (1984) who found no ABA in leaves of the
barley plastid (albino) mutant aibostrians. All dicot ABA-deficient mutants
are green and contain carotenoids. However, it seems unlikely that there is a
different ABA pathway in monocots and dicots. The occurrence of non-dor-
mancy and carotenoid deficiency in a sunflower mutant (Wallace and
Habermann, 1958) indicates that this relationship at least is not unique for
monocots. Most probably the absence of carotenoid mutants in dicots is due
to the fact that non-dormancy in dicots rarely expresses itself as vivipary.
When it does, it is less obvious because the viviparous seeds are covered by
the fruit tissue. Furthermore, lethal albinos in general are not maintained in
mutant collections.
Abscisic Acid 43
Carotenoid Intermediates Mutant alleles

phytoene
t vp2, vp5
phytofluene
t w3
~-carotene
t vp9
neurosporene
t
()-, y-carotene
t ps(= vp7)
a-, ~-carotene
t al (= y3)
a-, ~-cryptoxanthin
t
zeaxanthin, lutein
Fig. 2. Carotenoid biosynthesis in maize and the metabolic blocks of different
viviparous mutants (after Fong et al., 1983)
The ghost (gh) mutant of tomato, which is a developmentally unstable

chlorophyll and fruit carotenoid mutation (Rick et aI., 1959), deserves more
attention with respect to ABA effects. In their initial report neither dis-
turbed water relations nor deviating seed germination were mentioned.
The green wilty mutants apparently represent, if the C-40 pathway
operates, mutations in the last part of the ABA pathway . N evo and Tal
(1973) showed that jlaeea contained slightly higher levels of xanthophylls
(including violaxanthin) than its wild type, Rheinlands Ruhm. Recently
Bowman et al. (1984) observed the accumulation of 2,7-dimethyl-2,4-octa-
dienedioic acid (ODA) in jlaeea, which they suggest derives from oxidative
cleavage of violaxanthin (Taylor, 1984). However, as pointed out by Neill
and Horgan (1985), there are problems in explaining why this compound
accumulates when the ABA pathway is blocked between violaxanthin and
ABA. Although not conclusive, the biochemistry of both the viviparous
maize and jlaeea tomato mutants suggests the involvement of the indirect
C-40 pathway. No detailed biochemistry has been applied to the other
mutants mentioned above.
The interrelations between the three tomato mutants were studied gen-
etically by Taylor and Tarr (1984). When studying double mutants, they
found that the double recessive not/not, jle/jle and not/not, sit/sit had a
much more severe phenotype than the most extreme mutant sit. However,
injle/jle, sit/sit the effects were not additive. Taylor and Tarr give two pos-
sible explanations for these effects:
1. Fie and sit are mutated in a pathway different but parallel to a
pathway in which not regulates one step.
2. Fie and sit code for different subunits of the same enzyme and not for
a different enzyme in the same pathway.
44 M. Koomneef
D. The Use ofABA Mutants in Water Relations Research

The rapid wilting of ABA-deficient mutants in several species is due to
greater stomatal conductance, which can be normalized by exogeneously
applied ABA (Tal, 1966; Imber and Tal, 1970; Tal and Nevo, 1973;
Bradford, 1983; Neill and Horgan, 1985). This results in excessive transpi-
ration and wilting because stomata of the mutant remain partially open
during periods of water stress and in darkness. On the other hand, the
stomatal conductance of the fic tomato (Bradford et al., 1983) and wilted
pea (Donkin et al., 1983) is reduced by increased CO2 pressure, increased
leaf-to-air vapour pressure gradient and decreased light intensity, showing
that the regulation of stomatal closure still functions. Since these environ-
mental factors do not act by inducing ABA synthesis, it appears that ABA
overrides the factors mentioned above when water deficit threatens
(Bradford et al., 1983), thus enabling stomata to close despite the strong
promotion of stomatal opening by, for example, light. Apart from effects
on stomatal opening, hydraulic conductance (osmotic potential of the root
exudate) is also reduced in the mutant and this effect is ABA restorable
(Tal and Nevo, 1973; Bradford, 1983). Although not studied in as much
detail as tomato and pea, the overall picture for the potato (Waggoner and
Simmonds, 1966; Quarrie, 1982) and Arabidopsis (Koornneef et al., 1982)
mutants seems similar.
E. The Use ofABA Mutants in Seed Physiology

Seeds as they mature enter a period of arrest when development ceases and
dehydration enSl.les. The dehydrated seed may be either dormant or
quiescent. A quiescent seed will immediately start to germinate when it is
rehydrated. Dormancy is a physiological or morphological state that re-
stricts germination completely or to a limited range of environmental con-
ditions. Dormancy must be relieved before germination can occur. Based
on correlative evidence only, a regulatory role for ABA has been suggested
in the termination of seed growth, the start of dehydration, the prevention
of precocious germination, the stimulation of food reserves, the induction
of the developmental arrest and the induction of primary dormancy
(Karssen, 1982; Black, 1983). In Arabidopsis it was shown that the
induction of dormancy during seed development depends on endogenous
ABA. Immature seeds of the wild type germinate if they are isolated
half-way through development whereas mature seeds will not germinate. In
ABA-deficient mutants the germination potential increases instead of
decreases towards maturity (Fig. 3, after Karssen et al., 1983). The ABA
content during seed development in both genotypes is also shown in Fig. 3.
Reciprocal crosses of wild type and ABA mutants point to a dual origin
of endogenous ABA in seeds (Table 3). The genotype of the mother plant
regulated a sharp rise in ABA content with a maximum at 10 days after pol-
lination (maternal ABA). The genotype of the embryo was responsible for a
second ABA fraction (embryonic ABA) which reached far lower levels than
Abscisic Acid 45
100 100
80
.E
a
~60
.c
CII
~ c:
~40 40g
Q
c:
"---------q
,AF \
I
c:
"
cr"';
\
\
\
"e
"-
\ CD
\
\
20 C)
b..
... < I
8 10 12 14 16 18
Time after pollination, da ys
Fig. 3. Changes with time after pollination in ABA content ( -) and in germination
(- - -) of isolated seeds of the Arabidopsis wild type (open symbols) and the aba3
mutant (closed symbols). Modified after Karssen et al. (1983), with permission
maternal ABA, but persisted for some time during maturation after the
maternal ABA had decreased. Induction of dormancy only occurred if
embryonic ABA was present and was independent of the presence of
maternal ABA.
Applied ABA mimicked maternal ABA in this respect: dormancy was
not induced in ABA-deficient lines sprayed with a 100 11M ABA solution
every four days during seed development (Karssen, 1982; Karssen et aI.,
1983). Thus dormancy induction occurs entirely within the embryo. To
induce dormancy ABA must be synthesized close to its site of action or
must be localized in a specific subcellular compartment not accessible to
ABA transported from maternal tissues or supplied exogenously. These
results add further weight to the criticism of the classic hormone concept
which includes transport of hormones to the sites of action (Karssen, 1982).
That ABA in seeds partially originates in the mother plant has been
shown in the work on maize discussed in section B above. Using transloca-
tions with heterochromatic B chromosomes, Robertson (1955) showed that
the viviparous character in the vpl, vp8, vp5 and ps mutants, of which the
latter two are ABA deficient, is determined by the genotype of the embryo
and not the endosperm.
As previously mentioned, lack of endogenous ABA does not always
result in viviparous germination in the fruit. In Arabidopsis the dehydrated
state of the seeds prevents germination of the ABA mutants (Karssen et ai.,
46 M. Koornneef
Table 3. ABA content and germination of immature (10 and 16 days) and mature (26
days) F1 seeds of different crosses with Arabidopsis wild type and aba mutant
Days after pollination

10 16 ripe
9 x 0 F ABA Germ. ABA Germ. Germ.

ng/g % ng/g %
j
%
Aba/Aba x aba/aba Aba/aba 598 39 59 29 0

aba/aba x Aba/Aba Aba/aba 35 44 36 49 0
aba/aba x aba/aba aba/aba 8 64 1 99 95
1983), as do osmotic constituents of the pericarp tissue in tomato (Groot et

aI., pers. comm.). Robichaud et al. (1980) consider precocious germination
in maize to be due to the absence of developmental arrest in the mutant
because embryo development is already completed. Thus it is not only dor-
mancy that is under ABA control.
The role of ABA on food reserve accumulation has not been studied
using mutants. It is interesting to note, however, that in spite of the gen-
erally poor vegetative growth of the mutants, seed development is com-
parable to wild type.
F. ABA Deficiency in Relation to Other Physiological Effects

None of the processes such as fruit development, senescence and espe-
cially root geotropism (pilet, 1982; Audus, 1983) which have been asso-
ciated with ABA are clearly affected in ABA mutants (Koornneef, unpu-
blished; Moore and Smith, 1985). This suggests either that ABA plays no
part or that the residual levels of ABA in the mutants (below detection
level in maize) are sufficient for these processes.
III. Mutants Affecting ABA Sensitivity
Differences in ABA sensitivity for growth inhibition (Sloger and Caldwell,

1970) and the inhibition of seed germination (Rudnicki et aI., 1973) are found
between soybean cultivars. Stable variants with an up to 10-fold decrease in
ABA sensitivity (resistance to 0.1 mM ABA) have been selected in suspension
cultures of haploid Nicotiana sylvestris (Wong and Sussex, 1980). Preliminary
experiments suggest enhanced ABA catabolism in some of these variants (J.
R. Wong, pers. comm.). In carrot suspension cultures ABA at 0.1 mgl- l does
not inhibit cell growth, but particular stages of somatic embryo development
are affected, for example, the formation of plantlets from torpedo stage
embryos is inhibited (Sung et al., 1984). Variants that developed into plantlets
at this ABA concentration were selected and have been characterized as put-
ative uptake mutants (C. Borkird and Z. M. Sung, pers. comm.).
Abscisic Acid 47
An elegant selection scheme aiming at the selection of tissue-specific

alterations in ABA sensitivity was developed by Ho et ai. (1980) for barley
endosperm. Endosperm half seeds are placed on top of starch- and gibber-
ellin-containing agar with the cut edge in contact with the agar surface.
Gibberellin induced a-amylase synthesis is inhibited by ABA. After incu-
bation the agar plates are stained with an IrKI solution to visualize a-
amylase activity as a colourless halo around enzyme-producing half seeds.
ABA-insensitive mutants were identified as seeds that produced enzyme on
media containing both gibberellin and ABA.
Selecting amongst the progeny of mutagen-treated seeds for seeds that
germinated in ABA concentrations inhibitory to the germination of the pa-
rental variety, Hayter and Allison (1976) selected barley mutants with a
higher diastatic power (a measure of the activity of amylase enzymes). The
authors suggest that the increased enzyme activity is due to higher gibber-
ellin levels which overcome ABA inhibition. This is not, therefore, an
example of selection for true ABA insensitivity.
In the fern Ceratopteris, where ABA retards growth and inhibits sexual
differentiation in gametophytes, there has been a preliminary report of
several types of ABA resistant mutants (Hickok, 1984).
Mutants of Arabidopsis were selected on 10 J.1M ABA that were 5-20
times less sensitive to ABA at the germination and seedling growth level
than the wild type (Koomneef et ai., 1984). The ABA-insensitive mutants
have mutations at three different loci and show similarities in phenotype
(Table 4) with the ABA-deficient mutant. The altered sensitivity to ABA is
possibly due to reduced availability of a "receptor", or reduced binding-
capacity of a "receptor" for ABA. The "receptor" may be an ABA-binding
protein (Homberg and Weiler, 1984), a membrane fraction (Hocking et ai.,
1978) or the ABA target itself, i. e. molecules that react with
Table 4. Some characteristics of ABA-insensitive mutants of Arabidopsis (abi) as

compared to the wild type and the ABA-deficient mutant (aba)
Sensitivity to
Seed ABA
Genotype exogenous Transpiration
dormancy content
ABA
Wild type + + + +
abi-3 + +
abi-J/abi-2 +
aba +
Character: +
ABA sensitivity normal reduced
Seed dormancy normal reduced
Transpiration normal excessive
ABA content normal reduced
48 M. Koornneef
ABA or the ABA-receptor complex to give a physiological response. The

decreased sensitivity of these Arabidopsis mutants to exogenous ABA could
be also due to reduced ABA uptake (which would not explain the changes
in water relations and seed dormancy) or an enchanced ABA metabolism.
Enhanced metabolism is also unlikely since ABA levels in developing
seeds of the mutants were comparable or somewhat higher than in the wild
type (Koornneef et ai., 1984).
The vpl (= vp) mutant of maize is also insensitive to the germination
inhibiting effect of ABA when grown in vitro (Smith et ai., 1978; Robi-
chaud et ai., 1980). The mutant is green, has normal maize carotenoids and
normal ABA levels. It has been suggested that vpl is defective in a seed-
specific ABA receptor (Smith et ai., 1978). However, the absence of
flavenoid pigments in its endosperm, which is not a characteristic of other
viviparous, ABA-deficient maize mutants, is difficult to explain by this
hypothesis. Dooner (1985) noticed that in addition to three enzymes of
flavenoid synthesis viz. phenylalanine ammonia lyase, chalcone synthetase
and UDPG-flavenoid glucosyl transferase, the enzymes glucose-6-phos-
phate dehydrogenase, catalase and alcohol dehydrogenase, which in the
wild type show a similar pronounced increase in the later stages of
aleurone development, were also deficient in vpl aleurone layers. These
data suggest that vpl may be a major regulatory gene responsible for
turning on a battery of structural genes at a certain time in development
(Dooner, 1985). The relationship between ABA insensitivity and vivipary
remains unexplained by this hypothesis, unless an ABA receptor gene is
one of the genes regulated by vpl.
IV. Genetic Differences in ABA Accumulation
Reduction of leaf water potential, ABA accumulation and stomatal closure

result in the avoidance of excessive water loss in periods of water shortage.
Genetic variation in ABA accumulation and in the effect of ABA upon the
induction and relief of drought stress has therefore attracted the interest of
plant breeders. The subject has been reviewed recently by Quarrie (1983;
1984).
The levels to which ABA accumulated after stress induction and
drought resistance have been compared in cultivars of wheat (Quarrie and
Jones, 1979), maize (Larque-Saavedra and Wain, 1976; Ilahi and D6rffling,
1982), millet (Henson et ai., 1981) and sorghum (Durley et aI., 1983). The
reationship between ABA accumulation and drought resistance differed
between species, probably due mainly to the complexity of the interaction
of ABA accumulation with drought resistance, which is often measured as
crop yield. In contrast to beneficial effects on water use, high ABA levels
may also have detrimental effects on photosynthesis (by decreased
stomatal conductance), pollen fertility, etc. Quarrie (1984) mentions three
factors that determine how endogenous ABA modifies yield: 1) the amount
of ABA produced in response to drought and the organs to which it is dis-
Abscisic Acid 49
tributed, 2) the extent to which individual processes are regulated by a par-

ticular endogenous ABA level, and 3) the cumulative effects of interactions
between these processes and other factors that determine crop yield. In any
case, differences in experimental procedures (eg. how stress was applied)
and lack of isogenenicity make interpretation of the results difficult. The
near-isogenic material differing in ABA accumulation which has now been
obtained in wheat (Quarrie, 1981; Innes et al., 1984) and rice (Henson et
al., 1985) will provide much better material for studying ABA effects on
yield.
Apart from genetic differences in the extent to which external stress will
stimulate ABA production, genetic differences in ABA accumulation have
been reported for cotton (Ibragimov et al., 1978), maize (Larque-Saavedra
and Wain, 1976), millet (Henson, 1984), rice (Henson, 1983) and wheat
(Quarrie, 1981). The differences were shown to be heritable in millet, rice
and wheat. Homozygosity was obtained for this character in millet and
wheat after limited inbreeding, which, together with the high heritability,
suggests the segregation of only a small number of genes in the crosses
studied. In neither millet nor wheat (Quarrie and Henson, 1982) was there
evidence for dominance or a maternal influence on ABA accumulation.
However, in maize one example of maternal inheritance has been reported
(Larque-Saavedra and Rodrigues, 1979; Quarrie, 1983), suggesting the
genetic involvement of organelles (most probably plastids).
The biochemical basis of the differences in ABA accumulation reported
above has not been extensively studied. Durley et al. (1983) consider the
lower levels of ABA and higher levels of PA and total ABA metabolite con-
centration in particular genotypes to be the result of a higher efficiency of
conversion of ABA to its metabolites. Mutants that overproduce ABA con-
stitutively have not been described. The tomato mutants blind (hI) and
lateral suppressor (Is), characterized by increased apical dominance, were
initially reported to contain high levels of ABA (Tucker, 1976), but Taylor
and Rossall (1982), using reverse-phase, high-performance, liquid chroma-
tography instead of a bioassay, found no increased ABA levels in the Is
mutant. When ABA, PA and DPA levels of a similar torosa-2 mutant were
compared to wild type (Mapelli and Rocchi, 1985) no clear differences
were detected.
Differences in ABA levels have also been associated with dwarfness
(Quarrie, 1983). In comparisons between different non-isogenic apple
genotypes, an association of dwarfness with low ABA levels was found in
two studies (Lee and Looney, 1977; Yadava and Lockard, 1977), and with
higher levels in a third study (Robitaille and Carlson, 1976). The differ-
ences were not large and levels of other hormones were also affected,
therefore a causal relationship is doubtful. The association observed
between ABA accumulation and the presence of Rht genes in wheat cul-
tivars suggested a causal relationship, but comparison of near-isogenic
lines differing in Rht genes showed no such effect (Quarrie, 1983).
50 M. Koornneef
V. Conclusions
ABA-deficient mutants have been identified in a number of species.

Common to all the mutants is an excessive water loss due to a lower
stomatal conductance which results in a wilty phenotype. Most (probably
all) mutants also have reduced seed dormancy. These characteristics, espe-
cially the effects on seed dormancy, allow direct selection for ABA-defi-
cient mutants. Mutations resulting in ABA insensitivity can be selected
directly on the basis of ABA resistance. Several different mechanisms can
give rise to ABA insensitivity.
Genetic variation in ABA accumulation has been found among cul-
tivars, especially in cereals. No direct mutant selection schemes have yet
been developed for ABA overproducers. For this type of mutation, which
may be based upon reduced catabolism of ABA, direct screening using a
rapid ABA assay, like an immunoassay, may be applicable.
Although potentially of great importance, there has so far been only
limited use of mutants in ABA biochemistry, although in physiological
studies mutants have been used successfully to analyse the role of ABA in
stomatal opening and seed dormancy.
Acknowledgements
I thank many colleagues for providing preprints and other unpublished

information and discussions. In particular I would like to mention Drs. R.
Horgan, C. M. Karssen, J. M. Mottley, J. D. Smith, S. A. Quarrie and J. A.
D. Zeevaart.
VI. References
Addicott, F. T., Carns, H. R., 1983: History and introduction. In: Addicott, F. T.
(ed.), Abscisic Acid, pp. 1-21. New York: Praeger Scientific.
Alldridge, N. A., 1964: Anomalous vessel elements in wilty dwarf tomato. Bot. Gaz .
.125, 138-142.
Audus, L. J., 1983: Abscisic acid in root growth and geotropism. In: Addicott, F. T.
(ed.), Abscisic Acid, pp. 421-477. New York: Praeger Scientific.
Black, M., 1983: Abscisic acid in seed germination and dormancy. In: Addicott, F.
T. (ed.), Abscisic Acid, pp. 331-363. New York: Praeger Scientific.
Bowman, W. R., Linforth, R. S. T., Rossall, S., Taylor, I. B., 1984: Accumulation of
an ABA analogue in the wilty tomato mutant, flacca. Biochem. Genet. 22,
369-377.
Bradford, K. J., 1983: Water relations and growth of the flacca tomato mutant in
relation to abscisic acid. Plant Physiol. 72, 251-255.
Bradford, K. J., Sharkey, T. D., Farquhar, G. D., 1983: Gas exchange, stomatal
behaviour and Sl3C values of the flacca tomato mutant in relation to abscisic
acid. Plant Physiol. 72, 245-250.
Brenner, M. L., Burr, B., Burr, F., 1977: Correlation of genetic vivipary in corn with
abscisic acid concentration. Plant Physiol. Suppl. 63, 36.
Abscisic Acid 51
Creelman, R. A., Zeevaart, J. A. D., 1984: Incorporation of oxygen into abscisic

acid and phaseic acid from molecular oxygen. Plant Physiol. 75, 166-169.
Cummings, D. P., Stuthman, D. D., Green, C. E., 1978: Morphological mutations
induced with ethyl methanesulphonate in oats. J. Hered. 69, 3-7.
Donkin, M. E., Wang, T. L., Martin, E. S., 1983: An investigation into the stomatal
behaviour of a wilty mutant of Pisum sativum. J. Exp. Bot. 34, 825-834.
Dooner, H. K., 1985: Viviparous-l mutation in maize conditions pleiotropic
enzyme deficiencies in the aleurone. Plant Physioi. 77,486-488.
Durley, R. C., Kanangara, T., Seetharama, N., Simpson, G. M., 1983: Drought
resistance of Sorghum bieolor 5. Genotypic differences in the concentrations of
free and conjugated abscisic, phaseic and indole-3-acetic acids in leaves of
field-grown drought-stressed plants. Can. J. Plant Sci. 63, 131-145.
Fong, F., Koehler, D. E., Smith, J. D., 1983: Fluridone induction ofvivipary during
maize seed development. In: Krueger, J. E., La Berge, D. E. (eds.), III. Int.
Symp. on Pre-harvest Sprouting in Cereals, pp. 188-196. Boulder, Colo.:
Westview Press.
Goldbach, H., Michael, G., 1976: Abscisic acid content of barley grains during ri-
pening as affected by temperature and variety. Crop Sci. 16, 787 -800.
Hayter, A. M., Allison, M. J., 1976: Breeding for high diastatic power. In: Gaul, H.
(ed.), Barley Genetics III Proc. of the 3rd Int. Barley Genet. Symp., Garching
1975, pp. 612-619. Munchen: Verlag Karl Thiemig.
Henson, I. E., 1983: Abscisic acid accumulation in detached leaves of rice (Oryza
sativa L.) in response to water stress: a correlation with leaf size. Ann. Bot. 52,
385-398.
Henson, I. E., 1984: The heritability of abscisic acid accumulation in water-stressed
leaves of pearl millet (Pennisetum amerieanum [L.] Leeke). Ann. Bot. 53,1-11.
Henson, I. E., Mahalakshmi, V., Bidinger, F. R., Alagarswamy, G., 1981: Genotypic
variation in pearl millet (Pennisetum amerieanum [L.] Leeke) in the ability to
accumulate abscisic acid in response to water stress. J. Exp. Bot. 32, 899-910.
Henson, I. E., Loresto, G. C., Chang, T. T., 1985: Drought tolerance: Production of
closely-related lines of rice differing in drought-induced abscisic acid accumu-
lation. Int. Rice Res. Newsl., 10, 12-13.
Hickok, L. G., 1984: Selection and analysis of ABA-resistant mutants in Cera to-
pteris. Am. J. Bot. 71, 141.
Ho, T. H. D., Shih, S. c., Kleinhofs, A., 1980: Screening for barley mutants with
altered hormone sensitivity in their aleurone layers. Plant Physioi. 66, 153 -157.
Hocking, A., Clapham, J., Catsell, K. J., 1978: Abscisic acid binding to subcellular
fractions from leaves of Viciafaba. Planta 138, 303-304.
Hornberg, C., Weiler, E. W., 1984: High-affinity binding sites for abscisic acid on
the plasmalemma of Vieiafaba guard cells. Nature (London) 310, 321-324.
Ibragimov, A. P., Igamberdyeva, Z. I., Saidova, S. A., 1978: Effect of moisture
stress on the level of abscisic acid in cotton leaves. Uzb. BioI. Zh. 4, 11-14.
Ilahi, I., Dorffling, K., 1982: Changes in abscisic acid and proline levels in maize
varieties of different drought resistance. Physiol. Plant. 55, 129-135.
Imber, D., Tal, M., 1970: Phenotypic reversion of jlaeea, a wilty mutant of tomato,
by abscisic acid. Science 169, 592-593.
Innes, P., Blackwell, R. D., Quarrie, S. A., 1984: Some effects of genetic variation in
drought-induced abscisic acid accumulation on the yield and water use of
spring wheat. J. Agric. Sci. 102,341-351.
Karssen, C. M., 1982: Role of endogenous hormones during seed development and
52 M. Koornneef
the onset of primary dormancy. In: Wareing, P. F. (ed.), Plant Growth Sub-
stances 1982, pp. 623-632. London: Academic Press.
Karssen, C. M., Brinkhorst-van der Swan, D. L. c., Breekland, A. E., Koornneef,
M., 1983: Induction of dormancy during seed development by endogenous
abscisic acid: studies on abscisic acid deficient genotypes of Arabidopsis
thaliana (L.) Heynh. Planta 157, 158-165.
Koornneef, M., 1981: The complex syndrome of ttg mutants. Arabidopsis. Inf. Servo
18,45-51.
Koornneef, M., van der Veen, 1. H., 1980: Induction and analysis of gibberellin-
sensitive mutants in Arabidopsis thaliana (L.) Heynh. Theor. Appl. Genet. 58,
257-263.
Koornneef, M., lorna, M. L., Brinkhorst-van der Swan, D. L. C., Karssen, C. M.,
1982: The isolation of abscisic acid (ABA)-deficient mutants by selection of
induced revertants in non-germinating gibberellin-sensitive lines of Arabidopsis
thaliana (L.) Heynh. Theor. Appl. Genet. 61, 385-393.
Koornneef, M., Reuling, G., Karssen, C. M., 1984: The isolation and characteri-
zation of abscisic acid insensitive mutants of Arabidopsis thaliana. Physiol.
Plant. 61, 377 -383.
Koornneef, M., Cone, 1. W., Karssen, C. M., Kendrick, R. E., van der Veen, 1. H.,
Zeevaart, 1. A D., 1985: Plant hormone and photoreceptor mutants in Arabid-
opsis and tomato. In: Freeling, M. (ed.), Plant Genetics (UCLA Symposia on
Molecular and Cellular Biology, New Series, Vol. 35), pp. 103-114. New York,
NY: Alan R. Liss. Inc.
Larque-Saavedra, A, Rodriguez, G. M. T., 1979: Maternal inheritance of abscisic
acid (ABA) in Zea mays L. In: Abstracts of the 10th Intern. Conf. on Plant
Growth Subst. Madison, Wisconsin, p. 23.
Larque-Saavedra, A., Wain, R L., 1976: Studies on plant growth-regulating sub-
stances. XLII. Abscisic acid as a genetic character related to drought tolerance.
Ann. Appl. BioI. 83, 291- 297.
Lee, 1. M., Looney, N. E., 1977: Abscisic acid levels and genetic compaction in
apple seedlings. Can. 1. Plant Sci. 57, 81-85.
Mapelli, S., Rocchi, P., 1985: Endogenous abscisic acid in torosa-2 mutant tomato
plant. Plant Cell Physiol. 26, 371-374.
Marx, G. A, 1976: "Wilty": a new gene of Pisum. Pisum Newsl. 8, 40-41.
Milborrow, B. V., 1983: Pathways to and from abscisic acid. In: Addicott, F. T.
(ed.), Abscisic Acid, pp. 79 -111. New York: Praeger Scientific.
Milborrow, B. V., 1984: Inhibitors. In: Wilkins (ed.), Advanced Plant Physiology,
pp. 76-110. London: Pitman.
Moore, R, Smith, 1. D., 1984: Growth, graviresponsiveness and abscisic acid
content of Zea mays seedlings treated with fluridone. Planta 162, 342-344.
Moore, R, Smith, 1. D., 1985: Graviresponsiveness and abscisic-acid content of
roots of carotenoid-deficient mutants of Zea mays L. Planta 164, 126-128.
Neill, S. 1., Horgan, R., 1985: Abscisic acid production and water relations in wilty
tomato mutants subjected to water deficiency. 1. Exp. Bot. 36, 1222-1231.
Nevo, Y., Tal, M., 1973: The metabolism of abscisic acid injlacca, a wilty mutant of
tomato. Biochem. Genet. 10,79-90.
Pilet, P. E., 1982: Abscisic acid, one of the endogenous growth inhibitors regulating
root gravireaction. In: Wareing, P. F. (ed.), Plant Growth Substances 1982, pp.
529-536. London: Academic Press.
Postlethwait, S. N., Nelson, O. E., 1957: A chronically wilted mutant of maize. Am.
1. Bot. 44, 628-633.
Abscisic Acid 53
Quarrie, S. A, 1981: Genetic variability and heritability of drought-induced

abscisic acid accumulation in spring wheat. Plant Cell Environ. 4, 147 -151.
Quarrie, S. A, 1982: Droopy: a wilty mutant of potato deficient in abscisic acid.
Plant Cell Environ. 5, 23-26.
Quarrie, S. A, 1983: Genetic differences in abscisic acid physiology and their po-
tential uses in agriculture. In: Addicott, F. T. (ed.), Abscisic Acid, pp. 365-419.
New York: Praeger Scientific.
Quarrie, S. A, 1984: Abscisic acid and drought resistance in crop plants. News
Bull. Br. Plant Growth Regulator Group 7, 1-15.
Quarrie, S. A, Henson, I. E., 1982: Biparental inheritance of drought-induced accu-
mulation of abscisic acid in wheat and pearl millet. Ann. Bot. 49, 265-268.
Quarrie, S. A, Jones, H. G., 1979: Genotypic variation in leaf water potential,
stomatal conductance and abscisic acid concentration in spring wheat subjected
to artificial drought stress. Ann. Bot. 44, 323-332.
Quarrie, S. A, Lister, P. G., 1984: Evidence of plastid control of abscisic acid accu-
mulation in barley (Hordeum vulgare L.). Z. Pflanzenphysiol. 114, 295-308.
Rick, C. M., Thompson, A E., Brauer, 0., 1959: Genetics and development of an
instable chlorophyll deficiency in Lycopersicon esculentum. Am. J. Bot. 46,
1-11.
Robertson, D. S., 1955: The genetics ofvivipary in maize. Genetics 40, 745-760.
Robertson, D. S., 1975: Survey of the albino and white-endosperm mutants of
maize. J. Hered. 66, 67-74.
Robichaud, C. S., Wong, J., Sussex, I. M., 1980: Control of in vitro growth of vivip-
arous embryo mutants of maize by abscisic acid. Dev. Genet. 1,325-330.
Robitaille, H. A, Carlson, R. F., 1976: Gibberellin and abscisic acid-like substances
and the regulation of apple shoot extension. J. Am. Soc. Hortic. Sci. 101,
388-392.
Rudnicki, R., Blumenfeld, A, Bukovac, M. J., 1973: Glycine seed germination: Dif-
ferential response to abscisic acid. Experientia 29, 231.
Simmonds, N. W., 1965: Mutant expression in diploid potatoes. Heredity 20,
65-72.
Sloger, c., Caldwell, B. E., 1970: Response of cultivars of soybean to synthetic
abscisic acid. Plant Physiol. 46, 634-635.
Smith, J. D., McDaniel, S., Lively, S., 1978: Regulation of embryo growth by
abscisic acid in vitro. Maize Genet. Coop. N ewsl. 52, 107 -108.
Smith, J. D., Fong, F., Magill, C. W., Herlick, S., 1983: Fluridone-induced pheno-
copies of vp-5 in Zea mays seed. Genetics Suppl. 104, 66.
Stubbe, H., 1957: Mutanten der Kulturtomate, Lycopersicon esculentum, Miller I.
Kulturpflanze 5, 190-220.
Stubbe, H., 1958: Mutanten der Kulturtomate, Lycopersicon esculentum, Miller II.
Kulturpflanze 6, 89 -115.
Stubbe, H., 1959: Mutanten der Kulturtomate, Lycopersicon esculentum, Miller III.
Kulturpflanze 7, 82-112.
Sung, Z. M., Fienberg, A., Chorneau, R., Borkird, c., Furner, I., Smith, J., Terzi, M.,
Loschiavo, F., Giuliano, G., Pitto, L., Nuti-Ronchi, V., 1984: Developmental
biology of embryogenesis from carrot culture. Plant Mol. BioI. Rep. 2, 3-14.
Tal, M., 1966: Abnormal stomatal behaviour in wilty mutants of tomato. Plant
Physiol. 41, 1387-1391.
Tal, M., Nevo, Y., 1973: Abnormal stomatal behaviour and root resistance, and hor-
monal imbalance in three wilty mutants of tomato. Biochem. Genet. 8,
291-300.
54 M. Koornneef
Tal, M., Eshel, A, Witztum, A, 1976: Abnormal stomatal behaviour and ion
imbalance in Capsicum scabrous diminutive. J. Exp. Bot. 27, 953-960.
Tal, M., Imber, D., Erez, A., Epstein, E., 1979: Abnormal stomatal behaviour and
hormonal imbalance in jlacca, a wilty mutant of tomato. Plant Physiol. 63,
1044-1048.
Tal, M., Witztum, A., Shifriss, C., 1974: Abnormal stomatal behaviour and leaf
anatomy in Capsicum annuum, scabrous diminutive, a wilty mutant of pepper.
Ann. Bot. 38, 983-988.
Taylor, I. B., 1984: Abnormalities of abscisic acid accumulation in tomato mutants.
In: Menhennet, R., Lawrence, D. K. (eds.), Biochemical Aspects of Synthetic
and Naturally Occurring Plant Growth Regulators, Monograph 11, pp. 73-90.
Wantage: British Plant Growth Regulator Group.
Taylor, I. B., Rossall, S., 1982: The genetic relationship between the tomato
mutants, jlacca and lateral suppressor, with reference to abscisic acid accumu-
lation. Planta 154, 1-5.
Taylor, I. B., Tarr, A. R., 1984: Phenotypic interactions between abscisic acid defi-
cient tomato mutants. Theor. Appl. Genet. 68, 115-119.
Tucker, D. J., 1976: Endogenous growth regulators in relation to side shoot devel-
opment in the tomato. New Phytol. 77, 561-568.
Uknes, S. J., Ho, T. H. D., 1984: Mode of action of abscisic acid in barley aleurone
layers. Plant Physiol. 75, 1126-1132.
Waggoner, P. E., Simmonds, N. W., 1966: Stomata and transpiration of droopy
potatoes. Plant Physiol. 41, 1268-1271.
Wallace, R. H., Habermann, H. M., 1958: Absence of seed dormancy in a white
mutant strain of Helianthus annuus L. Plant Physiol. 33, 252-254.
Walton, D. C., 1980: Biochemistry and physiology of abscisic acid. Annu. Rev.
Plant Physiol. 31,453-489.
Wang, T. L., Donkin, M. E., Martin, E. S., 1984: The physiology of a wilty pea:
abscisic acid production under water stress. J. Exp. Bot. 35, 1222-1232.
Wong, J. R., Sussex, I. M., 1980: Isolation of abscisic acid-resistant variants from
tobacco cell cultures. II. Selection and characterization of variants. Planta 148,
103-107.
Yavada, U. L., Lockard, R. G., 1977: Abscisic acid and gibberellin in three
ungrafted apple (Malus sylvestris) rootstock clones. Physiol. Plant. 40, 225-229.
Zeevaart, J. A D., 1980: Changes in the levels of abscisic acid and its metabolites in
excised leaf blades of Xanthium strumarium during and after water stress. Plant
Physiol. 66, 672-678.
Chapter 3
Mutants as Tools for the Elucidation of Photosynthetic

Processes
Christa Critchley and Warwick Bottomley
Botany Department, The Faculties, Australian National University, GPO Box 4,

Canberra ACT 2601, and CSIRO Division of Plant Industry, GPO Box 1600, Can-
berra ACT 2601, Australia
With 4 Figures
Contents
I. Introduction
II. Two Genomes Code for the Structural and Regulatory Elements of the Photo-
synthetic Apparatus
III. Identification of Thylakoid Membrane Proteins and Their Functions
IV. From Phenotype to Gene Structure
A. Herbicide Resistance
B. Barley
C. Chlamydomonas
D. Arabidopsis
E. Denothera
F. Summary
V. Conclusions
VI. References
I. Introduction
Genetic, biochemical and physiological analyses of photosynthetic mutants

have been used for many years in attempts to broaden our understanding
of the structure-function relationships of the photosynthetic apparatus.
The recent development of molecular biology has given us some insight
into the primary structure of plant genes and is providing the techniques
for the investigation of the molecular basis of mutations which were previ-
ously only detectable through their phenotypic expression. In spite of these
advances we still have little understanding of the mechanisms regulating
the expression of these genes. The use of recombinant DNA techniques in
56 C. Critchley and W. Bottomley
directly probing and manipulating the molecular pathways responsible for

particular aspects of plant performance, such as photosynthesis, may lead
to significant advances in our knowledge and understanding of these pro-
cesses.
In this chapter we will neither deal with the genetic analysis of "photo-
synthetic mutants" as such nor will we provide a comprehensive review of
the recent literature on "photosynthetic mutants". We will instead attempt
a critical evaluation of the usefulness of these mutants to assist in the
solution of some of the physiological and biochemical problems con-
cerning the structure, organization and function of the photosynthetic
apparatus. We will draw on results of selected investigations which, we
believe, illustrate the power or otherwise of this particular approach. We
will also explore the potential for recombinant DNA technology to aid in
this approach by delineating the limitations of each of the disciplines
involved, i. e. genetics, molecular biology, biophysics, biochemistry and
physiology, stressing the importance of a multidisciplinary approach.
Chloroplasts, the organelles in the plant cells responsible for photosyn-
thesis, are bounded by a double membrane, the envelope, and contain a
highly organized and complex internal membrane system, the thylakoids,
in which the initial energy conversion and electron transport processes
occur. The fixation of carbon dioxide and the ensuing metabolic reactions
which lead to carbohydrate synthesis take place in the stroma, which is the
non-membranous part of the chloroplast. Photorespiration, part of which
F,
stroma
thy1ako.d
membrane
lumen
Pholosyslem n Cytochrome
Complex
b4' Photosystem I FerredoXIn
NADP - Reduclasc
Alp SYlllhase
Fig. 1. Model of the thylakoid membrane with the four major supramolecular com-
plexes designated Photosystem II, Photosystem I, Cytochrome b6/f complex and
ATP synthase. The numbers indicating molecular weights of individual polypep-
tides are those most commonly obtained from sodium dodecyl sulfatepolyacryl-
amide gel electrophoresis, which differ somewhat in most instances from those
obtained by deduction from gene sequences (for comparison see Table I)
Photosynthesis Mutants 57
also takes place in the stroma, is a process by which ribulose bisphosphate,

the acceptor for CO2 in photosynthesis, is oxidized by molecular oxygen.
The carbon reduction and oxidation cycles are interlocked and are both
initiated by the bifunctional enzyme ribulose-l,5-bisphosphate carbox-
ylase/oxygenase. The enzymatic reactions which constitute the two cycles
are well characterized biochemically and use the reducing power of the
NADPH 2 and the ATP generated in the membrane dependent light reac-
tions. Although most of the soluble enzymes involved in the carbon
reduction and oxidation cycles are well characterized, little is known of the
genetic factors which govern the regulation of expression and activity of
these enzymes. Much research effort has recently been concentrated on the
structure-function relationships of the constituents of the thylakoid mem-
brane and their organization.
Photosynthetic mutants have the potential to yield several categories
13
RuSP RuSP
"1 -t12
RSP XuSP
E4 P XuSP
bPGA
7
-.".,
FbP~GbP==<""GIP ,.v
bATP:xbHP 1
2T"-bADP bPi /
Hp.JIsi /
FBP Storch bDPGA
bPi~(bNADPH+HX+- 302
~
G3P DHAP G3P DHAP G3P
3r' bNADP
bG3P
b Hp
DHAP ~ \ \4 \ *4 \ )
/
Product
~Fe'ldbOCk -.-J
Fig. 2. The Reductive Pentose Phosphate Pathway (or Benson-Calvin Cycle) with
numbered stars indicating the enzymes involved: 1, Ribulose Bisphosphate Carbox-
ylase Oxygenase; 2, Phosphoglycerate Kinase; 3, Triosephosphate Dehydrogenase;
4, Triosephosphate Isomerase; 5, Aldolase (FBP Aldolase); 6, Fructose Bisphos-
phatase (FBPase); 7, Transketolase; 8, Aldolase (SBP Aldolase); 9, Sedoheptulose
Bisphosphatase (SBPase); 10, Transketo1ase; 11, Ribose Phosphate Isomerase; 12,
Ribulose Phosphate 3-Epimerase; 13, Phosphoribulokinase.
(With permission from Walker, D. A. and Edwards, G. C3, C4: mechanisms, and
cellular and environmental regulation of photosynthesis. Blackwell Scientific Publi-
cations, Oxford 1983)
of data. These include the identification of specific proteins which have

roles in photosynthetic processes, such as electron transport and carbon
metabolism, as well as the localization of such proteins in the supramo-
lecular complexes of the thylakoid membrane (Fig. I) or the positions of
enzymes in a metabolic pathway (Fig. 2). They can also be used to facilitate
biochemical and biophysical measurements when, for instance, the two
photo systems (Fig. 3) interfere with one another, as is the case when study-
ing ESR signals from photosystem II. Here, the use of a mutant deficient in
photosystem I components and function can eliminate interference from
this complex without the need to use detergents which may remove other
parts of the electron transport chain (see also Section IV).
In most of the past work on photosynthetic mutants neither the
molecular basis for the mutation(s) nor their location, i. e. whether chromo-
somal or extrachromosomal, was known because only in plants with which
formal genetic analysis was possible could it be established whether Men-
delian or maternal inheritance patterns prevailed. Therefore mutant work
was restricted to the detection of phenotypic expression of a lesion at the
biochemical level. Identification of a molecular defect or mutation, i. e. a
DNA-based alteration in gene structure, has only been possible since the
development of recombinant DNA techniques which enable positive iden-
tification of genes and their products, and the determination of their
-1 4
-1 2
-1 0 [pee 0'
-0
-0
\ Ph.o
'0
"0
-0 4
-0 2
PS II
I
> 0
.... +0 . 2
+0 . 4
+0
+0
.,

0
+1 2
Fig. 3. The Z-Scheme of photosynthetic electron transport indicating the redox

components operating within and between the two photosystems. SO"", S-states of
the oxygen evolving system; Z, secondary donor to PS II; P680, reaction centre of
PS II; Pheo, pheophytin, primary electron acceptor of PS II; Fe QA, quinone, first
stable electron acceptor of PS II; QB, quinone, secondary acceptor of PS II; PQ,
plastoquinone pool; FeS, Rieske iron-sulphur centre; Cyt f, cytochrome f; PC, plas-
tocyanin; P700, reaction centre of PS I; A 1, A2, primary and secondary electron
acceptors of PS I; Fd, ferredoxin
primary structure through DNA sequencing. Such a molecular defect or

change, however, needs to be correlated with, or shown to bear a causal
relationship to, the phenotypic characteristic in question. This causal rela-
tionship has to be extended further to the change in gene structure and the
physical manifestation of the mutation, such as alterations in the primary,
secondary or tertiary structure of the protein which may, for instance,
affect its binding affinity for substrates or cofactors or alter its temperature
dependence.
As long as the pleiotropic effects of a mutation are not ubiquitous
and the mutant characteristics can be described adequately and compre-
hensively by comparison with the wild type, mutants should be useful in
solving some of the biochemical problems in metabolism or membrane
structure. The question really is whether we can find examples showing
that the mutant approach has clearly contributed to our knowledge of pho-
tosynthetic membrane structure and function or metabolic pathways
which straightforward biochemical or biophysical analysis could not
provide.
II. Two Genomes Code for the Structural and Regulatory Elements
of the Photosynthetic Apparatus
It has long been known that chloroplasts contain their own DNA as well as
possessing the ability to express the genetic information encoded in that
DNA. They contain their own RNA polymerase and also ribosomes which
Table 1. Genes known to be located on the chloroplast genome
Gene Gene Product Complex Molecular

Weight
atpA CF1-alpha A TP-synthase 56,000

atpB CF1-beta A TP-synthase 54,000
atpE CF1-epsilon ATP-synthase 14,700
atpF CFo-1 ATP-synthase 24,000
atpH CFo-III ATP-synthase 8,000
rbcL large subunit RuBPC'ase 54,000
psaA reaction centre PS I 68,000
psbA QB-protein PS II 39,000
psbB reaction centre PS II 56,000
psbC reaction centre PSII 51,000
psbD D2 PSII 39,000
psbE cytochrome b 559 PS II 9,000
petA cytochrome f cytb6/f 33,000
petB cytochrome b 6 cytb6/f 23,000
petD polypeptide IV cytb6/f 17,000
are distinct from those of the cytoplasm (for reviews see Bohnert et aI., 1982;
Bottomley and Bohnert, 1982; Whitfeld and Bottomley, 1983). It is also well
documented that the genes coding for the chloroplast components are not
all contained in chloroplast DNA but, indeed, that most of these genes are
localized in the nucleus and are transcribed and translated in the cytoplasm.
Only a relatively small number of genes is coded for by chloroplast DNA.
Table I lists the genes currently known to be contained in chloroplast DNA.
Fig. 4 shows the physical map of the spinach chloroplast genome denoting
the approximate locations of the genes identified to date.
Most of the chloroplast polypeptides whose genes have been localized
occur in macromolecular complexes and it appears that the genetic infor-
mation for the constituents of each of these complexes is distributed
between the two genomes. That is, part of each complex is coded for by the
nuclear genome and part by the chloroplast DNA.
Fig. 4. A representation of the circular spinach chloroplast genome showing the

location of the known genes. The genes coding for constituents of the photosyn-
thetic system are as shown in Table I. Other genes are rps 19 and rpl2 - the genes
for the ribosomal proteins S 19 and L2; rRNA - the genes for the 23 S, 16 Sand 5 S
ribosomal RNAs
In addition, each chloroplast has, on average, about 25 copies of its

DNA and, since there are around 100 chloroplasts in each cell, it follows
that there are, on average, over 2000 copies of chloroplast DNA in each
green cell. While the mechanism of chloroplast DNA replication is largely
unknown at present, it is obvious that the means by which chloroplast
mutants arise is unlikely to be simple. A further complication is that the
nuclear encoded chloroplast genes that have been studied so far, those for
the small subunit of ribulose-l,5-bisphosphate carboxylase (Dunsmuir
et al., 1983 a; Coruzzi et al., 1983; Smith et al., 1983) and the chlorophyll
alb binding proteins of the light-harvesting complex of photo system II
(Dunsmuir et al., 1983 b; Dunsmuir, 1985; Cashmore, 1984), occur in mul-
tiple copies in each genome. These copies are not all identical but can be
divided into families according to the hybridizability of the untranslated
regions of the genes. These factors not only increase the complexity of the
interpretation of results derived from the use of mutants in the elucidation
of structural and functional aspects of photosynthesis but also of the iso-
lation of mutants with particular genotypic alterations.
III. Identification of Thylakoid Membrane Proteins and Their Functions
Mutants can facilitate biochemical and biophysical analysis of specific

photosynthetic functions (Moller et al., 1980; Nugent et al., 1980). Criteria
for assessment of photosynthetic properties are important because some
structural or functional characteristics may not be useful in mutant charac-
terization (Simpson and von Wettstein, 1980). Characterization of a photo-
synthetic mutant based on the presence or absence of components of a par-
ticular complex or the measurement of functional or biophysical properties
does not necessarily constitute an accurate characterization of the mutation
since they measure only phenotypic manifestations. An example of this is
the barley chlorina /2 mutant where the thylakoid membranes have been
shown to be deficient in chlorophyll-binding proteins of the photo system
II light-harvesting complex. However, it has also been shown that not only
are these proteins synthesized in vivo, but also that the chloroplasts of both
the wild type and the mutants are capable of transporting them across the
chloroplast envelope (Bellemare et al., 1982; Ryrie, 1983). This indicates
that the lesion affects assembly of the complex rather than synthesis of the
polypeptide. The mutant is known also to be deficient in chlorophyll band
it has been proposed by Bellemare et al. (1982) that normally the polypep-
tides are stabilized by binding to chlorophyll and the absence of this
binding results in an increased turnover of the proteins.
A good deal of work has been directed towards the identification of
polypeptides which are absent from mutant thylakoids. However, without
direct evidence concerning the synthesis of these peptides in the mutant
little can be concluded regarding the nature of the lesion.
Metz et al. (1980) have recently identified a 34 kD protein in PS II of
Scenedesmus as a component of the water splitting system, possibly
involved in manganese binding and maintenance of cytochrome b 559 HP.

It appears that this protein, in contrast to the 16 and 23 kD polypeptides of
the water splitting system (Honberg, 1984), is encoded in the chloroplast
DNA and it may well be identical with the recently sequenced so-called D2
polypeptide (Rasmussen et al., 1984; Holschuh et aI., 1984).
The Scenedesmus mutant LF-l, blocked on the oxidizing side of PS II
and shown to be deficient in a 34 kD protein, contained instead a 36 kD
protein (Metz and Seibert, 1984). These authors also confirmed that this
34 kD protein is not identical with the 33 kD "Murata" protein involved in
oxygen evolution (Miyao und Murata, 1984; Kuwabara et aI., 1985).
Recently Browse et al. (1985) and McCourt et al. (1985) described a
mutant of Arabidopsis thaliana which is deficient in the unusual fatty acid
trans-3-hexadecenoic acid. This acyl group exclusively esterifies the
2-position of phosphatidylglycerol (PG), one of the phospholipid consti-
tuents of normal thylakoid membranes, and has been thought to play an
important role in the structure and/or function of these membranes. Signif-
icantly, the complete lack of this fatty acid does not alter the structural or
functional characteristics of the mutant chloroplasts when compared to
those of the wild type, either by electron microscopy of grana structure or
by measurement of photosynthetic activities for PS I and PS II partial reac-
tions in vitro. Electrophoretic analysis of the chlorophyll-protein com-
plexes and properties of chlorophyll fluorescence under various conditions
showed no significant effect on energy transfer efficiency or related prop-
erties in the mutant (McCourt et al., 1985). This is an example of a useful
photosynthetic mutant which has assisted in the elucidation of photosyn-
thetic structure and function. Indeed it has been shown very elegantly that
this unusual fatty acid, the subject of much controversy concerning its role
in the membrane, is not required for either structure or function in the thyl-
akoids.
IV. From Phenotype to Gene Structure
A. Herbicide Resistance
In 1984, the first reports appeared of a mutation affecting a protein
involved in photosynthesis for which it is believed the molecular alteration
in the primary gene structure is known. This is a mutation of the rapidly
turned over QB membrane protein that is an intrinsic component of PS II
and is thought to bind the herbicides atrazine and diuron. Interestingly, the
gene for this protein is encoded in the chloroplast DNA and is extremely
highly conserved between a number of diverse plant species. The only dif-
ference in the amino-acid sequences, as deduced from the nucleotide
sequences of the genes of the sensitive and resistant biotypes, is a substi-
tution for the serine at position 264 in the protein sequence as a result of a
single base mutation. In Solanum nigrum (Hirschberg et al., 1983; Golou-
binoff and Edelman, 1984) and Amaranthus hybridus (Hirschberg and
Mcintosh, 1983) glycine is the substitute while in Chlamydomonas it is

alanine (Erickson et al., 1984). The coincidence of triazine resistance and a
single amino-acid substitution in the triazine binding protein provides only
circumstantial evidence, although quite compelling and appealing evidence
nonetheless, that the substitution is the direct cause of the alteration in the
affinity of the polypeptide for triazine. The future challenge is to elucidate
the nature of the alteration in the secondary or tertiary structure of the
protein which will explain the change in triazine-binding properties
leading to resistance, and perhaps to introduce the suspected specific
mutation into the genome of a suseptible species and thereby confer tri-
azine resistance on its photosynthetic electron transport function.
B. Barley
A great deal of painstaking work has been carried out, mainly by the
Carlsberg Laboratories in Copenhagen, on a series of barley mutants with
alterations resulting in various differences in their photosynthetic capabil-
ities. The mutations have mainly been characterized in terms of fluores-
cence properties, photosystem activities and ultramicroscopic structure.
One such mutant which has a potential to localize genes for chloroplast
constituents in either the nuclear or the chloroplast DNA has been
described by Hoyer-Hansen and Casadoro (1982). This mutant is a condi-
tional temperature sensitive mutant of barley which, if grown at the per-
missive temperature of 30 C, contains complete sets of functional ribo-
somes whereas, if grown at 23 C, appears to lack both 70 S ribosomes and
chloroplast ribosomal proteins. Examination of the polypeptide compo-
sition of the chloroplast thylakoids showed that nine polypeptides known
to be synthesized on chloroplast ribosomes were missing.
However, it has also been shown that the mutation is not in any of the
genes associated with ribosome assembly or function but rather in the syn-
thesis of carotenoids. The destruction of ribosomes at the lower temperature
is a secondary effect of the mutation and can be overcome by selection of the
wavelength of light in which the plants are grown. At long wavelengths even
at 23 C there is a normal complement of chloroplast ribosomes. This
mutant has been used to show that the gene for a 23 kD component of the
water-splitting complex associated with photo system II is located in the
nucleus, since the protein is present in the mutant when grown under the
conditions where the chloroplast ribosomes are absent (Honberg, 1984).
This 23 kD component of the water-splitting complex is also present in
the membranes of mutants which are deficient in photosystem II. Thus it
appears that the water-splitting complex is capable of assembling in the
thylakoid membrane in the absence of a complete photo system II complex.
Hoyer-Hansen et al. (1979) found that three mutants deficient in stroma
thylakoids also have decreased amounts of all five subunits of the eF1
complex of ATP-synthase. This finding supports the contention that the
ATP-synthase complex is localized exclusively on the stroma thylakoids
and does not occur within the granal stacks.
While some of the mutants appear to be deficient in a single protein,

others, such as some of the photo system I mutants, have deficiencies in at
least four polypeptides. This is probably due to pleiotropic effects such as
occur in the chlorina /2 mutant which contains decreased amounts of four
photo system I proteins and three proteins of the photo system II light-har-
vesting complex. However, it has been demonstrated by Bellemare et al.
(1982) that all of these proteins are synthesized in the plants and that they
are transportable across the chloroplast membrane. Although a number of
mutants in PS II which are deficient in only one polypeptide are known, all
PS I mutants so far isolated are deficient in at least three polypeptides and
it would be useful if PS I mutants lacking individual polypeptides were
available (Hoyer-Hansen et al., 1982).
Moller et al. (1980), Nugent et al. (1980) and Moller et al. (1981)
describe some of the biochemical properties of photosystem I mutant
viridis-n 34. The thylakoid membranes of this mutant are deficient in four
polypeptides, one of which is the chlorophyll-protein-l of PS I and the
other three polypeptides are believed to be iron-sulphur proteins. All four
proteins are found in PS I particles isolated from the wild type.
EPR measurements on PS I particles from this barley mutant show that
the characteristic signals for PS I are present in those particles which
contain only three polypeptides of 110, 18.3 and 15.2 kD (Moller et al.,
1981).
EPR measurements of PS II signals are facilitated by the use of PS I
mutant viridis-b 63, because of the lack of interference from PS I signals.
These studies supplement and support those with PS II particles and whole
thylakoids.
Haworth et al. (1982) claim that barley mutant chlorina /2 lacks the
LHC II polypeptide which is phosphorylated in vivo and in vitro and which,
through phosphorylation and dephosphorylation, regulates the excitation
energy distribution between PS II and PS I. They claim that this lack of
protein and phosphorylation make state 1 - state 2 transitions impossible.
By contrast, Ryrie (1983) showed that three of the four LHC II poly-
peptides are present in chlorina /2, although in greatly reduced quantities,
and that phosphorylation also takes place. This would appear to be an
instance where mutant work has confused rather than clarified the issue in
question and we do not seem to have learned anything about either the role
of LHC II in excitation energy distribution or that of its phosphorylation.
The lack of mutants with specific and definable lesions together with the
difficulties associated with pleiotropic effects, complicate the interpre-
tation of results obtained with these mutants.
C. Chlamydomonas
The work on a variety of Chlamydomonas mutants designed to unravel the
relationship between photosynthesis, bicarbonate transport and carbonic
anhydrase activity, for example, has generated a lot of data that are quite
difficult to interpret. It has, however, unambiguously shown that the co or-
dinated regulation of all three processes is necessary, and has provided a

direct demonstration that a component other than carbonic anhydrase is
involved in the carbon dioxide concentrating mechanism (Spalding et ai.,
1983; Ogren et ai., 1984; Spalding et ai., 1985).
D. Arabidopsis
Somerville (1984) has recently published an excellent review article on the
subject of carbon fixation and photorespiration mutants. We will therefore
only deal with one more recent example. The dicarboxylate transporter
mutant of Arabidopsis thaliana is an example of how biochemical and phy-
siological analysis of a mutant helped to determine the specificity of the
transporter and its primary in vivo role, which were either previously not
known or disputed (Somerville and Ogren, 1983). Subsequently Somerville
and Somerville (1985) described the comparison of chloroplast envelope
polypeptides of wild type and transporter mutant and found that a 42 kD
protein was missing from the polypeptide pattern of the mutant envelope.
This protein is suggested to be a component of the dicarboxylate trans-
porter, about whose structural identity nothing was known previously.
E. Oenothera
The chloroplast encoded genes of most plants used in the study of photo-
synthesis, indeed of most higher plants, exhibit maternal inheritance. That
is, the total genome of the F 1 plastids is derived from the female parent.
Segregation is in the ratio of 0: 4 instead of the usual ratio of 1 : 2: 1, which
is the Mendelian ratio found in nuclear inheritance, where equal contribu-
tions to the genetic make-up come from each parent. One genus in which
chloroplasts can be derived from either parent is Oenothera. In this genus
the chloroplast genomes of either parent can be transferred to the progeny.
An extensive study of the inheritance patterns of the genus has been made
by Stubbe and Herrmann (1982) who have described five types of chloro-
plast genomes and three types of nuclear genomes. Stubbe and his co-
workers described the phenotypic effects of transferring the various chloro-
plast types into the three types of nuclear background, particularly with
regard to the chlorophyll content (Kutzelnigg and Stubbe, 1974). They iso-
lated more than 40 of these mutants varying from green to white and
described some combinations which were lethal.
Unfortunately, only a relatively small amount of biochemical work has
been carried out on these mutants to date and only one instance of a
location of a mutation has been reported. The mutant, named Isigma, has
been shown to lack the large subunit of ribulose-1,5-bisphosphate carbox-
ylase (Hallier et al., 1978). More recent work (Hildebrandt et al., 1984),
using immunoblotting of polyacrylamide gels, has revealed that the mutant
produces a smaller polypeptide reacting with antiserum to the large
subunit, suggesting that the gene contains a mutation which causes the
premature termination of"transcription or translation.
This mutant has also been used in an effort to shed some light on the
mechanism of coordination of the expression of chloroplast and nuclear
genes. Since the gene for the large subunit of ribulose-l,5-bisphosphate
carboxylase is located in the chloroplast DNA, while that for the small
subunit is nuclear encoded, and they are required in equal numbers in the
enzyme complex, the possibility exists that the presence or absence of one
of the subunits affects the synthesis of the other. For instance Ellis et al.
(1978) have suggested that the synthesis of the small subunit regulates the
synthesis of the large subunit. One way to test this hypothesis would be to
determine whether the inhibition of the synthesis of one subunit affected
the production of the other. If the small subunit was controlling the syn-
thesis of the large subunit then, under conditions where synthesis of the
large subunit was inhibited by some means other than by inhibiting the
synthesis of the small subunit, it could be expected that the small subunit
would still be present or even would accumulate in amounts in excess of
normal. A mutant in which one subunit is not produced is obviously ideal
material for such an experiment. Hallier et al. (1978) used this lsigma Oeno-
thera mutant in such an experiment and came to the conclusion that, on the
basis of polyacrylamide gels and immunodiffusion, no small subunit was
produced by the mutant. This was in contrast with the results of Feier-
abend (1978) who used a temperature-sensitive mutant of rye which con-
tained no chloroplast ribosomes and therefore synthesized no large subunit
when grown at the nonpermissive temperature. This author found that the
small subunit was still produced in the absence of the large subunit. More
recently the lsigma mutant has been re-examined by Hildebrandt et al.
(1984), who found that the small subunit was indeed produced in the
mutant. However, since they also showed that a truncated large subunit
was produced by the mutant the situation is by no means clear and the pos-
sibility remains that the presence of the small subunit is affecting the syn-
thesis of the large subunit.
Sears and Herrmann (1985) have examined an Oenothera mutant defi-
cient in two of the eight subunits of the A TP-synthase complex. Although
the results of the in vivo experiments suggested that the two subunits, which
in all plants so far examined are juxtaposed on the genome and cotran-
scribed (Whitfeld and Bottomley, 1983), are fused to one another perhaps
by a deletion of the adjacent ends, translation of the mRNA from this
mutant yielded polypeptides which were indistinguishable from those of
the wild type. Sears and Herrmann (1985), therefore, suggest that the
mutation is in some component which affects a post-transcriptional event.
This then can presumably also be classed as a pleiotropic effect which
results in an apparent photosynthetic mutation.
F. Summary
Genes which code for proteins involved in the photosynthetic process are
distributed between the nuclear and the chloroplast genomes. Of the genes
coding for proteins which have so far been shown to occur in chloroplast
DNA, the majority have been genes for proteins that are involved directly
in the photosynthetic process. That is, with the exception of genes for
tRNAs, ribosomal proteins (Schmidt et al., 1983), elongation factor, ribo-
somal RNA and RNA polymerase, the genes so far located are either con-
stituents of the photosynthetic membranes or the large subunit of ribu-
10se-l,5-bisphosphate carboxylase (see Table 1 and Fig.4). The number of
structural proteins whose genes are located on the chloroplast genome is
small compared to the total number of genes required for the development
and function of the chloroplast. The significance of the occurrence of those
specific genes may become more apparent when one or several complete
chloroplast genomes have been sequenced and analyzed. Another remark-
able aspect of the distribution of genes for proteins associated with photo-
synthesis is that, to date, no incidence has been reported of differences in
the location of a gene for a specific protein; in other words at present there
is an absolute commonality of the distribution of genes between the
nucleus and plastid among and across plant species from Euglena and
Chlamydomonas to higher plants. It is most interesting in this context to
note that the genes which code for the small and the large subunits of ribu-
lose bisphosphate carboxylase, which in all higher plants and algae with
chloroplasts are genetically separated, are linked in the blue-green alga
Anabaena, which has only one genome and no separate plastid-bound thyl-
akoid membranes. The two subunits are cotranscribed so that the
requirement for coordinated expression would appear to be less of a
problem in these primitive organisms (Nierzwicki-Bauer et al., 1984).
It seems to be fairly well established now that there is no mRNA
exchange between the two sites of protein synthesis: this is to say that all
cytoplasmic mRNA appears to be translated on 80 S cytoplasmic ribo-
somes, whereas all chloroplastic mRNA seems to be translated on 70 S
chloroplast ribosomes. Chloroplast mRNA is generally believed not to be
polyadenylated and this structural feature may serve to distinguish
between sites of protein synthesis, and may in fact be related to structural
differences between 70 Sand 80 S ribosomes. Another feature of the dis-
tinction between the two genomes is the finding that most nuclear-encoded
proteins which are destined for the chloroplast are synthesized as pre-
cursors, whereas most chloroplast-coded proteins are not. The leader
sequences on the nuclear-encoded proteins are involved in their transport
across the chloroplast membrane, since these sequences are cleaved off
during passage into the organelle. As pointed out before, the complexity
associated with the dual genomes clearly increases the complexities and
the ambiguity associated with interpretation of the results obtained with
mutants.
It was highlighted above that, considering the number of chloroplasts
per cell and the number of DNA copies per chloroplast, each cell should
contain around 2000 copies of chloroplast DNA. To date, only one
organism has been reported to contain heterogeneous chloroplast DNA
molecules (Jenni et al., 1981). It was found that the mUltiple copies of the
Euglena gracilis chloroplast genome are not uniform in size and that the
difference between the extremes amounted to about 800 basepairs or

somewhat less than 0.6 % of the average genome length. These authors
could not distinguish whether the differences were intra- or inter-chloro-
plastic or -cellular. It is, however, quite feasible that such differences, if
they were intrachloroplastic for example, could be the equivalent of the
multigene families found for nuclear genes (Dunsmuir et al., 1983 a, b;
Cashmore, 1984; Broglie et al., 1983). Both phenomena may have a role in
the differential expression of a particular gene in response to environ-
mental or other regulatory factors.
The greater majority of plant chloroplast genomes so far investigated
contain an inverted repeat sequence of around 20 kbp, which suggests the
possibility that the DNA circle can exist in two forms as a result of intra-
molecular recombination at the ends of these repeats. This was indeed
shown to be so for the two cases in which a restriction enzyme could be
found which did not cut within the repeats, thus allowing the detection of
two different sized fragments containing the inverted repeats. These cases
were Phaseolus (Palmer, 1983) and also the cyanelle DNA from Cyano-
phora paradoxa (Bohnert and Loeffelhardt, 1982).
V. Conclusions
The photosynthetic process is a complex one carried out in a specialized

organelle with the aid of a number of protein complexes in a highly struc-
tured membrane system and a large amount of soluble enzymes in the
stroma. In addition, the genetic information coding for these components
is contained in at least two distinct genetic systems. The expression of these
two genomes is carried out by two different transcription and translation
systems, one of which, that of the chloroplast, is largely procaryotic in
nature while that of the cytosol is eucaryotic. In view of the complexity it is
obvious that mutants should be extremely useful tools for the elucidation
of the mechanisms of the photosynthetic processes as well as that of the
regulation and coordination of the two systems. There is no doubt that
mutants have played, and are continuing to play, an important role in fur-
thering our understanding of these processes. In particular they can
provide material in which the complexity is reduced and interpretation is
simplified by the removal of parts of the systems. However, the overall pro-
cesses are so complex that the interpretation of the results is still difficult,
particularly because of the possibility of phenotypic expression being the
result of indirect interactions of the constituents, as has been demonstrated
a number of times.
To date, the contributions of recombinant DNA technology to the
solution of the problems of photosynthetic mechanisms are only relatively
minor. The potential for the application of such techniques as site directed
mutagenesis is just beginning to be tapped. For example, the elegant work
of Gruissem et al. (1983) who have used this technique, combined with in
vitro transcription in a chloroplast cell-free system, to localize the regu-
latory regions of the gene for tRNAmet and the work of Mullet et ale (1985),
who applied the same approach to the genes for the large subunit of ribu-
lose bisphosphate carboxylase and the beta subunit of ATP-synthase, give
an indication of the advances in our understanding of the subject that
should be made in the near future.
The application of the knowledge gained from the understanding of the
photosynthetic processes to the solution of problems in plant science will
ultimately depend on the availability of a transformation system for chloro-
plasts. While there are indications that this will soon be achieved, the
problem remains as to the possible complications caused by the existence
of the large number of copies of chloroplast DNA in each cell.
Some recent research that may have implications on the interpretation
of results using mutants are the reports that copies of parts of the chloro-
plast DNA are present in other genomes in the cell. It was been reported
by Lonsdale et ale (1983) that maize mitochondrial DNA contains regions
homologous with both chloroplast ribosomal cistrons as well as the gene
for the large subunit of ribulose bisphosphate carboxylase. In addition,
Scott and collaborators (Scott and Timmis, 1984; Whisson and Scott, 1985)
have reported that the nuclear DNA of spinach contains regions homol-
ogous with chloroplast DNA. Further substantiation and extension of
these results, together with the possible demonstration that these extra
chloroplast DNA genes are expressed in vitro, would not only complicate
the interpretation of some results but would also raise interesting questions
regarding the mechanism of such duplications and their possible role in the
overall regulation of photosynthesis.
VI. References
Bellemare, G., Bartlett, S. G., Chua, N.-H., 1982: Biosynthesis of chlorophyll a/b-
binding polypeptides in wild type and the chlorina f2 mutant of barley. J. BioI.
Chem. 257, 7762-7767.
Bohnert, H.-J., Crouse, E. J., Schmitt, J. M., 1982: Organization and expression of
plastid genomes. In: Parthier, B., Boulter, D. (eds.), Nucleic Acids and Proteins
in Plants II. Encyclopedia of Plant Physiology, New Series, Vol. 14 B, pp.
475-530. New York - Berlin - Heidelberg: Springer-Verlag.
Bohnert, H.-J., Lieffelhardt, W., 1982: Cyanelle DNA from Cyanophora paradoxa
exists in two forms due to intramolecular recombination. FEBS Lett. 150,
403-406.
Bottomley, W., Bohnert, H.-J., 1982: The biosynthesis of chloroplast proteins. In:
Parthier, B., Boulter, D. (eds.), Nucleic Acids and Proteins in Plants II. Ency-
clopedia of Plant Physiology, New Series, Vol. 14B, pp. 531-596. New York-
Berlin - Heidelberg: Springer-Verlag.
Broglie, R., Coruzzi, G., Lamppa, G., Keith, B., Chua, N.-H., 1983: Structural
analysis of nuclear genes coding for the precursor to the small subunit of wheat
ribulose-l,5-bisphosphate carboxylase. Bio/Technology 1, 55-61.
Browse, J., McCourt, C. R., Somerville, C. R. 1985: A mutant of Arabidopsis lacking
a chloroplast-specific lipid. Science 227, 763-765.
Cashmore, A. R., 1984: Structure and expression of a pea nuclear gene encoding a
chlorophyll alb-binding polypeptide. Proc. Nat. Acad. Sci., U.S.A. 81,
2960-2964.
Coruzzi, G., Broglie, R, Cashmore, A., Chua, N.-H., 1983: Nucleotide sequence of
two pea cDNA clones encoding the small subunit of ribulose 1,5-bisphosphate
carboxylase and the major chlorophyll alb-binding thylakoid polypeptide. 1.
BioI. Chem. 258, 1399-1402.
Dunsmuir, P., 1985: The petunia chlorophyll alb-binding protein genes: a com-
parison of Cab genes from different gene families. Nuc. Acids Res. 13,
2503-2519.
Dunsmuir, P., Smith, S. M., Bedbrook, 1., 1983 a: A number of different nuclear
genes for the small subunit of RuBCase are transcribed in petunia. Nuc. Acids
Res. 11,4177-4183.
Dunsmuir, P., Smith, S. M., Bedbrook, 1., 1983 b: The major chlorophyll a/b-
binding protein of Petunia is composed of several polypeptides encoded by a
number of distinct nuclear genes. J. Mol. Appl. Genet. 2, 285-300.
Ellis, R 1., Highfield, P. E., Silverthorne, 1., 1978: The synthesis of chloroplast pro-
teins by subcellular systems. In: Hall, D.O., Coombs, 1., Goodwin, T. W. (eds.),
Photosynthesis 1977, pp. 497-506.
Erickson, 1. M., Rahire, M., Bennoun, P., De1epelaire, P., Diner, B., Rochaix, J.-D.,
1984: Herbicide resitance in Chlamydomonas reinhardii results from a mutation
in the chloroplast gene for the 32-kilodalton protein of photosystem II. Proc.
Nat. Acad. Sci., U.S.A. 81, 3617-3621.
Feierabend,l., 1978: Cooperation of cytoplasmic and plastidic protein synthesis in
rye leaves. In: Akoyunoglou, G., Argyroudi-Akoyunoglou, 1. H. (eds.), Chloro-
plast Development, pp. 207-213. Amsterdam: Elsevier.
Goloubinoff, P., and Edelman, M., 1984: Chloroplast-coded atrazine resistance in
Solanum nigrum: psbA loci from susceptible and resistant biotypes are isogenic
except for a single codon change. Nuc. Acids Res. 12,9489-9496.
Gruissem, W., Greenberg, B. M., Zurawski, G., Prescott, D. M., Hallick, R. B.,
1983: Biosynthesis of chloroplast transfer RNA in a spinach chloroplast tran-
scription system. Cell 35,815-828.
Hallier, U. W., Schmitt, 1. M., Heber, U., Chaianova, S. S., Volodsarsky, A. D.,
1978: Ribulose-5-bisphosphate carboxylase-deficient p1astome mutants of Oen-
othera. Biochim. Biophys. Acta 504, 67 -83.
Haworth, P., Kyle, D. 1., Arntzen, C. 1., 1982: Protein phosphorylation and excit-
ation energy distribution in normal, intermittent-light-grown, and a chlorophyll
b-1ess mutant of barley. Arch. Biochem. Biophys. 218, 199-206.
Hildebrandt, 1., Bottomley, W., Moser, 1., Herrmann, R G., 1984: A plastome
mutant of Oenothera hookeri has a lesion in the gene for the large subunit of
Ribulose-l,5-bisphosphate carboxylase/oxygenase. Biochim. Biophys. Acta
783,67-73.
Hirschberg, 1., Bleecker, A., Kyle, D. 1., McIntosh, L., 1983: The molecular basis of
triazine-herbicide resistance in higher-plant chloroplasts. Z. Naturforsch. 39c,
412-420.
Hirschberg, 1., McIntosh, L., 1983: Molecular basis of herbicide resistance in Ama-
ran thus hybridus. Science 222, 1346-1349.
Holschuh, K., Bottomley, W., Whitfeld, P. R, 1984: Structure of the spinach chlor-
oplast genes for the D 2 and 44 kD reaction-centre proteins of photosystem II
and for tRNASer (UGA). Nuc. Acids Res. 12,8819-8834.
Honberg, L. S., 1984: Probing barley mutants with a monoclonal antibody to a
polypeptide involved in photosynthetic oxygen evolution. Carlsberg Res.

Commun. 49, 703-719.
Hoyer-Hansen, G., Moller, B. L., Pan, L. c., 1979: Identification of coupling factor
subunits in thylakoid polypeptide patterns of wild-type and mutant barley thy-
lakoids using crossed immunoelectrophoresis. Carlsberg Res. Commun. 44,
337-351.
Hoyer-Hansen, G., Casadoro, G., 1982: Unstable chloroplast ribosomes in the cold-
sensitive barley mutant tigrina-034 Carlsberg Res. Commun. 47, 103-118.
Hoyer-Hansen, G., Moller, B. L., Henrey, L. E. A., Casadoro, G., 1982: Thylakoid
polypeptide synthesis and assembly in wild-type and mutant barley. Cell
Function and Differentiation, Part B, 111-125.
Jenni, B., Fasnacht, M., Stutz, E., 1981: The multiple copies of the Euglena gracilis
chloroplast genome are not uniform in size. FEBS Lett. 125, 175-179.
Kutzelnigg, H., Stubbe, W., 1974: Investigations on plastome mutants in Oenothera.
1. General considerations. Sub-Cell. Biochim. 3, 73-89.
Kuwabara, T., Miyao, M., Murata, T., Murata, N., 1985: The function of 33-kDa
protein in the photosynthetic oxygen-evolution system studied by reconstitution
experiments. Biochem. Biophys. Acta 806, 283-289.
Lonsdale, D. M., Hodge, T. P., Howe, C. J., Stern, D. B., 1983: Maize mitochon-
drial DNA contains a sequence homologous to the ribulose-l,5-bisphophate
carboxylase large subunit gene of chloroplast DNA. Cell 34, 1007-1014.
McCourt, P., Browse, J., Watson, J., Arntzen, C. J., Somerville, C. R, 1985:
Analysis of photosynthetic antenna function in a mutant of Arabidopsis thaliana
(L.) lacking trans-hexadecenoic acid. Plant Physiol. 78, 853-858.
Metz, J. G., Seibert, M., 1984: Presence in photosystem II core complexes of a
34-kilodalton polypeptide required for water photolysis. Plant Physiol. 76,
829-832.
Metz, J. G., Wong, J., Bishop, N. I., 1980: Changes in electrophoretic mobility of
chloroplast membrane polypeptide associated with the loss of the oxidizing
side of photo system II in low fluorescent mutants of Scenedesmus. FEBS Lett.
144,61-66.
Miyao, M., Murata, N., 1984: Effect of urea on photo system II particles: Evidence
for an essential role of the 33 kilodalton polypeptide in photosynthetic oxygen
evolution. Biochim. Biophys. Acta 765, 253-257.
Moller, B. L., Nugent, J. H. A., Evans, M. C. W., 1981: Electron paramagnetic res-
onance spectrometry of photo system I mutants in barley. Carlsberg Res.
Commun. 46, 373-382.
Moller, B. L., Smillie, R. M., Hoyer-Hansen, G., 1980: A photo system I mutant in
barley (Hordeum vulgare L.). Carlsberg Res. Commun. 45, 87 -99.
Mullet, J. E., Orozco, E. M., Chua, N.-H., 1985: Multiple transcripts for higher
plant rbcL and atpB genes and localization of the transcription initiation site of
the rbcL gene. Plant Mol. BioI. 4, 39-54.
Nierzwicki-Bauer, S. A., Curtis, S. E., Haselkorn, R, 1984: Cotranscription of genes
encoding the small and large subunits of ribulose-l ,5-bisphosphate carboxylase
in the cyanobacterium Anabaena 7120. Proc. Nat. Acid Sci., U.S.A. 81,
5962-5965.
Nugent, J. H. A., Moller, B. L., Evans, M. C. W., 1980: EPR detection of the
primary photochemistry of photo system II in a barley mutant lacking photo-
system I activity. FEBS Lett. 121, 355-357.
Ogren, W. L., Somerville, C. R, Somerville, S. c., Spreitzer, R J., Spalding, M. H.,
Jordan, D. B., 1984: Genetic analysis of photosynthetic carbon pathways. In:
Sybesma, C. (ed.), Advances in Photosynthesis Research, Vol. III, pp. 429-435.

The Hague: NijhoffiJunk.
Palmer, J. D., 1983: Chloroplast DNA exists in two orientations. Nature (London)
301,92-93.
Rasmussen, O. F., Bookjans, G., Stummann, B. M., Henningsen, K. W., 1984:
Localization and nucleotide sequence of the gene for the membrane poly-
peptide D 2 from pea chloroplast DNA. Plant Mol. BioI. 3, 191-199.
Ryrie, I. J., 1983: Immunological evidence for apoproteins of the light-harvesting
chlorophyll-protein complex in a mutant of barley lacking chlorophyll b. Eur. J.
Biochem. 131, 149-155.
Schmidt, R. J., Richardson, C. B., Gillham, N. W., Boynton, J. E., 1983: Sites of
synthesis of chloroplast ribosomal proteins in Chlamydomonas. J. Cell BioI. 96,
1451-1463.
Scott, N. S., Timmis, J. N., 1984: Homologies between nuclear and plastid DNA in
spinach. Theor. Appl. Genet. 67, 279-288.
Sears, B. B., Herrmann, R. G., 1985: Plastome mutation affecting the chloroplast
ATP synthase involves a post-transcriptional defect. Current Genet. 9,
521-528.
Simpson, D. J., von Wettstein, D., 1980: Macromolecular physiology of plastids.
XIV. Viridis mutants in barley: genetic, fluoroscopic and ultrastructural charac-
terization. Carlsberg Res. Commun. 45, 283-314.
Smith, S. M., Bedbrook, J., Spiers, J., 1983: Characterisation of three cDNA clones
encoding different mRNAs for the precursor to the small subunit of wheat ribu-
losebisphosphate carboxylase. Nuc. Acids Res. 11,8719-8734.
Somerville, C. R., 1984: The analysis of photosynthetic carbon diox.ide fixation and
photorespiration by mutant selection. Oxford Surveys of Plant Molecular &
Cell Biology 1, 103 -131.
Somerville, S. C., Ogren, W. L., 1983: An Arabidopsis thaliana mutant defective in
chloroplast dicarboxylate transport. Proc. Nat. Acad. Sci., U.S.A. 80,
1290-1294.
Somerville, S. C. Somerville, C. R., 1985: A mutant of Arabidopsis deficient in chlor-
oplast dicarboxylate transport is missing an envelope protein. Plant Sci. Lett.
37,217-220.
Spalding, M. H., Spreitzer, R. J., Ogren, W. L., 1983: Genetic and physiological
analysis of the CO 2-concentrating system of Chlamydomonas reinhardii. Planta
159,261-266.
Spalding, M. H., Spreitzer, R. H. Ogren, W. L. 1985: Use of mutants in analysis of
the COz-concentrating pathway of Chlamydomonas reinhardii. In: Lucas, W. J.,
Berry, J. A., (eds.), Inorganic carbon uptake by aquatic photosynthetic
organisms. American Society of Plant Physiology, Rockville, MD, pp.
361-375.
Stubbe, W., Herrmann, R. G., 1982: Selection and maintenance of plastome
mutants and interspecific genome/plastome hybrids from Oenothera. In:
Edelman, M., Hallick, R. 8., Chua, N.-H. (eds.), Methods in Chloroplast
Molecular Biology, pp. 149-165. Amsterdam: Elsevier.
Whisson, D. L., Scott, N. S., 1985: Nuclear and mitochondrial DNA have sequence
homology with a chloroplast gene. Plant Mol. BioI. 4, 267 -273.
Whitfeld, P. R., Bottomley, W., 1983: Organization and structure of chloroplast
genes. Annu. Rev. Plant Physiol. 34, 279-310.
Chapter 4
Maize Alcohol Dehydrogenase: A Molecular Perspective
Wayne L. Gerlach, Martin M. Sachs*, Danny Llewellyn,

E. Jean Finnegan and Elizabeth S. Dennis
Division of Plant Industry, CSIRO, GPO Box 1600, Canberra, ACT 2601, Australia
*Dept. of Biology, Washington University, St. Louis, MO 63130, U.S.A.
With 5 Figures
Contents
I. Introduction
II. Genetics and Expression of ADH Enzymes in Maize
A. Two Genes Encode ADH in Maize
B. Organ Specificities of ADHI and ADH2 Activities
C. Mutations of Adh Genes of Maize
D. The Maize Anaerobic Response
III. Isolation of Adh Genes
A. cDNAs from Anaerobically Induced Maize Genes
B. Isolation and Identification of Adhl cDNA Clones
C. Isolation and Identification of Adh2 cDNA Clones
D. Isolation of Adh Genes from Other Plant Species
IV. Structure of Plant Adh Genes
V. Three-Dimensional Structure of ADH Enzymes
VI. Genetic Change Around and Within Adh Genes of Maize
A. Allelic Variation
B. Somaclonal Variation
C. Gene Duplication
D. Ds Element Mutations in Adhl
E. Robertson's Mutator
VII. Approaches to the Mechanism of Adh Gene Regulation
A. DNA Sequence Comparisons
B. Mutations in Adh Which Affect Expression
VIII. In Vivo Expression of Adh
A. Attempts at Expression in Cereal Tissue Culture Cells
B. Testing Maize Adh Gene Activity in Dicotyledonous Plant Cells
C. Expression of Maize Adh in Animal Cells
IX. Conclusions
X. References
74 W. L. Gerlach et al.
I. Introduction
During the 1960's a number of advances were made toward understanding

the regulation of prokaryotic and fungal gene systems. From there interest
turned to gene regulation in higher eukaryotes and, amongst plant gene
systems, the genetics and biochemistry of the maize alcohol dehydroge-
nases (ADH, Ee 1.1.1.1) received immediate attention from two groups
(Scandalios, 1966; Schwartz, 1966; Schwartz and Endo, 1966).
There were several reasons for this interest in the maize ADH system.
Maize itself was readily amenable to genetic analysis and a strong back-
ground knowledge of maize genetics already existed. ADH was found to be
very stable in extracts and could be isolated to homogeneity (Felder et al.,
1973; Kelly and Freeling, 1980). Analysis of ADH activity was fairly
straightforward using a spectrophotometric assay (Hageman and Flesher,
1960; Freeling, 1973). An activity-specific stain was also developed and
this could be applied to both gel electrophoresis and in situ studies of cells
and tissues (Schwartz and Endo, 1966). Furthermore, studies had shown
that ADH activity could be induced by flooding maize seedlings (Hageman
and Flesher, 1960). If, as was thought, this induction of activity acted at the
gene level then the ADH system seemed to provide an opportunity to
approach the study of gene expression and its regulation in a plant system.
The alcohol dehydrogenase system of maize has now been extensively
characterized in genetic, physiological and, more recently, molecular terms.
This chapter will review this work from the characterization of the system
in terms of enzyme activity and how this led to the isolation of the genes,
through the molecular organization of the Adh genes and on to the insights
this has provided about this and other genetic systems of maize. Because of
the large body of knowledge which has accumulated on the Adh genes of
maize, we will primarily confine this review to the maize ADH system, but
will also draw upon information about Adh genes of other plants whenever
relevant to the concepts under consideration.
II. Genetics and Expression of ADH Enzymes in Maize
A. Two Genes Encode ADH in Maize

Gel electrophoresis of inbred maize lines reveals three ADH isozymes.
Genetic analysis using different inbred lines (Schwartz and Endo, 1966)
and EMS-induced mutant lines (Schwartz, 1969 a, 1969 b) which exhibit
simultaneous polymorphisms in electrophoretic migration or disap-
pearance of two of the three isozymes showed that ADH activity is
encoded by two unlinked genes (Freeling and Schwartz, 1973). These are
Adh 1 on chromosome 1 (Schwartz, 1971) and Adh 2 on chromosome 4
(Dlouhy, 1980). Additionally, the ADH enzyme is active as a dimer and the
three isozymes are explained as ADHl.ADHI and ADH2.ADH2 homo-
dimers and the ADHl.ADH2 heterodimer (Freeling, 1974).
Alcohol Dehydrogenase 75
B. Organ Specificities ofADHI and ADH2 Activities

Adhl and Adh2 were found to be expressed differently relative to each
other in different maize tissues and organs. Neither ADHI nor ADH2 is
found to any great extent in seedling tissues, but both activities are induced
by flooding and by treatment with the synthetic plant hormone
2,4-dichlorophenoxyacetic acid, (2,4-D) (Freeling, 1973). There are high
levels of ADHI and ADH2 in the naturally developing endosperm but
little or no activity in mature endosperm. ADH1 activity is predominant in
the scutellum, aleurone, embryo and anther, while ADH2 predominates in
tassel nodes and the peduncle. ADH1 is found in pollen but there is no
detectable ADH2 activity.
C. Mutations ofAdh Genes of Maize

This topic has been extensively reviewed by Free1ing and Birchler (1981)
and more recently by Freeling and Bennett (1985) and this section will,
therefore, give only a brief indication of the mutational analyses of maize
Adh genes.
Numerous natural variants of Adhl have been characterized. These
include variants that affect electrophoretic mobility (e. g. S, F, C; Schwartz
and Endo, 1966), specific activity (e. g. c m and natural null alleles;
Schwartz and Laughner, 1969; Stuber and Goodman, 1983), ability or
inability of alleles to intragenically recombine with each other (Freeling,
1976, 1978), relative organ specific expression (Woodman and Freeling,
1981), transcript length variants (Dennis et al., 1984; Sachs et al., 1986),
polymorphisms in regions flanking the Adhl gene (Johns et al., 1983; Sachs
et al., 1985) and a natural, tightly linked duplication of Adhl (Fc rn ;
Schwartz and Endo, 1966).
A number of variants of Adh2 have been characterized by Dlouhy
(1980; described also in Freeling and Birchler, 1981). These include electro-
phoretic variants (R, P, N, L) and naturally occurring null alleles. Adh2
alleles also exhibit differences in relative expression in the presence of eth-
ylene (Schwartz, 1978), and extensive restriction site polymorphisms
(Dennis et al., 1985).
Three methods have been developed to screen for induced Adhl
mutants. The allozyme screen, developed by Schwartz (see Freeling and
Birchler, 1981) involves mutagenesis of one Adhl electrophoretic allele and
crossing to a different electrophoretic variant. Ml seeds are then analysed
for mutation. Allyl alcohol selection of pollen (Schwartz and Osterman,
1976; Freeling and Cheng, 1978) takes advantage of the huge number of
pollen grains from a single plant and the relative insensitivity of pollen to
allyl alcohol unless it has normal levels of ADH which convert the alcohol
to the highly toxic aldehyde, acrolein. A third method can be used to select
in both directions, for and against ADHI activity. When kernels are
flooded with aerated water, seeds with normal levels of ADH1 will ger-
minate. Those with lower levels will not germinate but can be subsequently
76 w. L. Gerlach et al.
rescued by germination in air (Chen, pers. comm; see also Freeling and
Bennett, 1985).
Several Adhl mutants have been isolated after EMS treatment, mostly
by Schwartz (see Freeling and Birchler, 1981; Freeling and Bennett, 1985).
Many, perhaps all, appear to be point mutations most easily explained as
base substitutions. They include mutants with altered electrophoretic
mobility, specific activity, heat lability, dimerization ability, as well as
CRM + and CRM - nulls. Most appear to have normal levels of normal
sized mRNAs and none appear to affect gene regulation. One exception is
S664 which appears to have an unusually long transcript in addition to one
of the normal size and seems to be affected in transcription or 3' processing
(Hake et ai., 1984). Most of the remaining Adhl mutants have been isolated
by insertion of transposable elements. These will be discussed in more
detail below.
D. The Maize Anaerobic Response

Hageman and Flesher (1960) were the first to show that ADH activity
increases as a result of flooding maize seedlings. Freeling (1973) later
showed that ADH activity increased at a zero order rate between 5 hr and
72 hr of anaerobic treatment reflecting a simultaneous expression of two
unlinked genes, Adhl and Adh2. Schwartz (1969a) showed that ADH
activity is required to allow maize seeds and seedlings to survive anaerobic
treatment and presumably flooding under natural conditions. Since this
anaerobic response formed the basis of the isolation of the Adhl and Adh2
genes we will consider it here in more detail.
In many ways the maize anaerobic response is analogous to the heat-
shock response described in many animals and plants. However, except for
one possible overlap, it involves a different set of proteins (Sachs et ai.,
1980; Cooper and Ho, 1983; Kelley and Freeling, 1982). There is an imme-
diate repression of pre-existing protein synthesis and the first five hours of
anaerobic treatment are a transition period during which there is a rapid
increase in the synthesis of a class of polypeptides with an approximate
molecular weight of 33 kd (Fig. 1). These have been referred to as the trans-
ition polypeptides (TPs). After approximately 90 minutes, the synthesis of
an additional group of 20 novel "anaerobic" polypeptides (ANPs) is
induced, amongst which are ADHI and ADH2. This group of 20 ANPs
represents more than 70 percent of the total label incorporation after five
hours of anaerobic treatment. By this time synthesis of the TPs is at a
minimal level; however, pulse-chase experiments have shown that the TPs
are very stable. The synthesis of the ANPs continues in a quantitatively
stable ratio for up to 72 hours of anaerobic treatment, at which time
protein-synthesis decreases as the seedlings begin to die (Sachs et ai., 1980).
The rapid repression of pre-existing protein synthesis, as seen also in
anaerobically treated soybean seedlings (Lin and Key, 1967), is correlated
with a near complete dissociation of polysomes in anaerobically treated
maize tissue (Dennis and Pryor, unpublished). It is not due to degradation
8
ro.JA - All .....
50s
,,'03d
+ 87- _
77-

~
()
.~
--........
.J o
. . II:
_. - _ ..
.. ~g= - P"
:- - 53 ,5 2..
45- - - ' . ti
(1)
42 ::::::-. P"
TP. TI' '<
o 0-
~'"
3
335-~ 0-
33 31.5
a
(JQ
(1)
;:l
III
(1)
'"
. ,J
- --.-. ~-- ----------
Fig. 1. Proteins synthesized in maize primary roots during aerobic (A) and anaerobic (B; 12-17 hours anaerobiosis)
growth. Proteins being synthesized were radioactively pulse labelled and visualized by autoradiography after two
dimensional electrophoresis. The unlabelled arrow points to ADH I location. TP arrow points to position of tran-
sition polypeptides (Sachs et al., 1980)
-.J
-.J
of "aerobic" mRNAs, since the mRNAs encoding the "aerobic" proteins

remain translatable in an in vitro system at least five hours after anaerobic
treatment is initiated (Sachs et al., 1980). This agrees with the observation
that in soybean seedlings the polysomes dissociated by anaerobiosis can
rapidly reform to 80-90 percent of their pretreatment levels, even in the
absence of new RNA synthesis, when soybean seedlings are returned to air
(Lin and Key, 1967).
The functions of some of the ANPs are known. ADHI and ADH2 have
been identified as ANPs through the use of genetic variants (Sachs and
Freeling, 1978; Ferl et al., 1979). More recently glucose phosphate isom-
erase and fructose-l,6-diphosphate aldolase have been identified as ANPs
(Kelley and Freeling, 1984a, b) and sucrose synthetase may also be an
ANP (P. Starlinger, M. Freeling, pers. comms.). Pyruvate decarboxylase
activity has also been shown to be induced by anaerobiosis (Wignarajah
and Greenway, 1976; Laszlo, 1981) and therefore may be an ANP. The
functions of the remaining ANPs and TPs are as yet unknown. However,
all five of the ANPs that have been identified are glycolytic enzymes. Thus,
it appears that at least one function of the anaerobic response is to
enable the plant to produce as much ATP as possible during short-term
flooding.
In the presence of air, each maize organ examined, including the roots,
coleoptile, mesocotyl, endosperm, scutellum and anther wall, synthesizes a
tissue-specific spectrum of polypeptides. Under anaerobic conditions all of
the above organs synthesize only the ANPs. Moreover, except for a few
characteristic qualitative and quantitative differences, the patterns of an-
aerobic protein synthesis in these diverse organs are remarkably similar
(Okimoto et al., 1980). On the other hand, maize leaves which have
emerged from the coleoptile do not incorporate label under anaerobic con-
ditions and do not survive even very short periods of anaerobiosis
(Okimoto et aI., 1980).
III. Isolation of Adh Genes
A. cDNAsfrom Anaerobically Induced Maize Genes

In vitro translation studies provided the basis for cloning the Adh genes of
maize. Fundamental was the observation that translatable Adhl and Adh2
mRNAs were more abundant in anaerobically treated seedlings than in
aerobically treated seedlings (Ferl et al., 1980). Therefore, a cDNA library
prepared from mRNA isolated from anaerobically treated maize seedlings
should contain cDNA sequences of the maize Adh genes.
RNA was isolated from anaerobically treated seedlings and frac-
tionated according to size by sucrose-gradient centrifugation. An RNA size
class which produced ADH-sized polypeptides in in vitro translation exper-
iments was chosen for cloning reactions. A cDNA library was prepared
from this RNA and transformant colonies were hybridized in parallel with
two different radioactive probes prepared from RNA of anaerobically or

aerobically grown plants. Colonies were chosen which showed greater
hybridization with the anaerobic probe than with the aerobic probe (Fig.
2A). These were presumed to contain cDNA sequences complementary to
induced transcripts in the anaerobic RNA population, amongst which
would likely be cDNAs from ADH mRNAs (Gerlach et al., 1982).
B. Isolation and Identification ofAdhl cDNA Clones
The chimaeric plasmids from individual clones were used to hybrid-select

mRNAs from an mRNA population prepared from anaerobic seedlings.
The selected mRNAs were then translated in vitro. One clone, pZML84,
selected an mRNA which translated to a 40 kd polypeptide, the same size
as ADHI and ADH2. Further tests, which drew upon the wealth of back-
ground information on maize Adh genetics, indicated that it was a cDNA
clone for Adhl of maize.
First of all, RNAs from genetic variants were used in hybrid-selection
experiments. One experiment involved RNA from a line which produced a
variant of ADH with altered electrophoretic mobility (the Adhl-U725
allele). It was found that the clone pZML84 selected an mRNA which
translated in vitro to produce an ADH polypeptide with the expected alter-
ation in electrophoretic migration. When RNA from a heterozygote for
alleles encoding ADHs with these different electrophoretic mobilities (gen-
otype: Adhl-F / Adhl-U725; Ferl et al., 1979) was used, the hybrid selection
experiment produced the two products expected.
Physical properties of the maize ADH1 polypeptide were also used to
verify that the pZML84 hybrid selection product was ADH1 polypeptide
and that the clone therefore contained an Adhl cDNA sequence. It was
found that the HRT product co-migrated with authentic ADH1 (Ferl et al.,
1979) on two dimensional isoelectric focussing/polyacrylamide gel electro-
phoresis. Previous techniques of purification of RNA had also led to pro-
duction of antibody against ADHI. A polyclonal antibody prepared
against ADH1 also recognized and precipitated the HRT product. Con-
clusive evidence came from a comparison of amino-acid sequences of
ADH1 protein and the amino-acid sequence predicted from the cDNA
nucleotide sequence (Dennis et al., 1984). Three fragments were isolated
from purified ADH1 protein after partial proteolytic digestion and sub-
jected to sequence analysis from their N-termini. The amino-acid
sequences were found in the inferred sequence from the cDNA clone, as
were other peptide sequences which had been determined in M. Freeling's
laboratory (Kelly and Freeling, unpublished).
In later experiments (Dennis et al., 1984) other Adhl cDNA clones were
isolated using the original pZML84 cDNA insert sequence as a probe. The
cDNA sequences in these newer clones had been size-fractionated prior to
cloning to ensure that only long cDNA sequences were included. Among
these, one clone designated pZML793 included the entire 3' non-coding
Adh1 Adh2
Fig. 2. Colony hybridizations. (A) Signals seen when duplicate filters containing
arrays of different anaerobic cDNA clones were hybridized with radioactive probes
prepared from anaerobic "An" or aerobic "Aer" mRNAs. Colony 84 contains
sequences corresponding to an anaerobically induced RNA and proved to be Adh 1
cDNA. (B) Different signal strengths seen when Adh 1 cDNA is used as probe on
colonies containing either homologous Adh 1 cDNA plasmid or heterologous Adh 2
cDNA (80 % sequence homology, see text)
region, the translated region and extended into the 5' non-coding region
i. e. a near full-length Adhl cDNA clone.
C. Isolation and Identification ofAdh2 cDNA Clones

The Adhl cDNA clone was used in the isolation of Adh2 cDNA sequences.
During experiments in which the Adhl cDNA sequence was used as a
probe to isolate further cDNA clones in colony hybridization experiments,
colonies were observed which hybridized to the probe under moderate
stringency, albeit at a lower level than colonies containing authentic Adhl
cDNA (Fig. 2B). Because ADH1 and ADH2 can form functional heterod-
imers and crossreact immunologically, it was considered likely that they
would be related in amino-acid sequence and, hence, that their genes might
have related nucleotide sequences. Thus, it seemed probable that these
weakly cross-hybridizing colonies contained Adh2 cDNA sequences.
Isolation and sequencing of these cDNA sequences showed that they
probably did encode Adh2 (Dennis et ai., 1985). The cDNA sequence was
82 % homologous with the Adhl sequence and derived amino-acid
sequences had 87 % identity. Verification that the clone contained Adh2
cDNA again used the existing genetic information on the maize ADH
system. It was known that lines containing the Adh2-33 null allele did not
accumulate translatable Adh2 mRNA as measured by in vitro translation
(Ferl et al., 1980) and the putative Adh2 cDNA did not hybridize to mRNA
from any line containing this allele, whereas it did hybridize to all lines
containing active Adh2 alleles.
Genomic clones containing gene sequences and flanking regions of
Adhl and Adh2 from maize have been isolated from libraries using these
cDNA probes (e. g. see Dennis et al., 1984, 1985).
D. Isolation ofAdh Genes from Other Plant Species

The cloned Adh sequences from maize can be used as probes to detect and
isolate cross-hybridizing sequences in cloned libraries from other plants.
Isolation and subsequent characterization, particularly sequence analysis,
of such clones will determine whether they do in fact contain Adh
sequences from the plant under consideration.
This approach has been used in our laboratory to isolate three Adh
cDNA clones from pea (E. J. F and D. L., manuscript in preparation). They
were detected by sequence homology with the maize Adhl cDNA
sequence. Sequence data have been obtained for all three clones and they
appear to be from two distinct but very similar Adhl-like genes with 96 %
homology over 386 bp of coding region and 89 % over 150 bp of 3'-untrans-
lated region. When compared with the nucleotide and predicted amino-
acid sequences of maize Adhl cDNA, the degree of homology is 78 % and
85 % respectively. When compared with maize Adh2, the values are 73 %
and 84 % respectively.
Cross hybridization with the maize Adhl cDNA was also used to isolate
an Adh genomic clone from barley (M. Trick and K. Edwards, pers.
comm.), and a 2.3 kb Hind III restriction fragment from within the Adhl
genomic sequence of maize has been used to isolate a cross-hybridizing
genomic clone from Arabidopsis thaliana (c. Chang and E. Meyerowitz,
pers. comm.). In both cases sequencing of the isolated clones indicates that
they encode ADH polypeptides.
The use of prokaryotic expression vectors has enabled the isolation of
an Adh cDNA clone from tomato cell suspension cultures (B. Williams,
pers. comm.) in which tomato ADH2 is found at high levels. A library of
cDNA clones inserted into the ~-galactosidase gene of a phage lambda
vector was prepared. Colonies which produced ADH polypeptide as a
~-gal fusion protein were detected by immunological cross reaction with
polyclonal and monoclonal antibodies to tomato ADH. The cloned tomato
cDNA sequences have homology with the maize Adh genes and are,
therefore, from an Adh gene. Again, the ability to isolate pure ADH for the
production of antibodies was central to the success of this approach.
IV. Structure of Plant Adh Genes
Genomic clones of maize Adhl-IS and Adh2-N alleles have been fully
sequenced and permit a comparison of the two genes (Fig. 3; and Dennis et
al., 1984, 1985). The coding regions appear to be identical in length and are
more than 80 % homologous at both the amino-acid and nucleotide
sequence level. Most nucleotide sequence differences occur in the
redundant third base positions of codons. Both genes are interrupted by
nine introns in identical positions within the gene. These introns account
for up to half of the length of the genes. Except for the short regions at
intron-exon boundaries which contain splice signals, there is very little
length or sequence resemblance between corresponding introns in Adhl
and Adh2. The splice signals are essentially the same as those found in
other plant and animal genes, reflecting the probable similarity in RNA
splicing mechanism between animals and plants. The close structural ho-
mology and similarity in number and location of introns in the genes is evi-
dence that they arose by duplication and subsequent divergence from a
single ancestral gene.
One interesting region of difference between AdhI and Adh2 lies in
exon 5. This is shown in Fig. 4 A. One interpretation is that compensating
frame-shift mutations have occurred in this region in one of the Adh genes
during their evolution. Presumably the non-essential nature of the Adh
genes (lines containing Adh-null mutations are viable under normal aerobic
growth conditions) has allowed the original frame shift null mutation to be
tolerated until a compensating frameshift mutation occurred nearby to
restore activity.
Another noteworthy structural rearrangement which has occurred
during evolution of these genes is found at the 3' end of intron 6 of the
AdhI gene (Dennis et al., 1984). Here a region of approximately 100 bp is
5' Noncoding Introns 3' Noncoding

2 3 4 5 6 7 8 9
Adhl 100 bp 535bp 97 419 86 94 342 87 91 90 300-400 bp
Adh2 126 95 101 235 95 124 495 125 90 112 315
Coenz~e Bindin~ Domain
Catalytic Domain
- - - - - - -
Fig. 3. Structural features of maize Adh 1-S and Adh2-N alleles. Each gene contains
9 introns at identical locations in the coding sequence. Intron sizes as well as 5'
leader and 3' non-coding region lengths are shown. Relationship of coding regions
to protein structure are shown on the lower lines
repeated three times (Fig. 4 B). The first two repeats are entirely within
intron 6. Both contain in-frame stop codons following the putative splice
acceptor sites, and so could only code for a truncated protein if they were
incorrectly used as the intron exit during splicing of an RNA transcript.
The third copy of the repeat spans the junction of intron 6 and its adjacent
exon region. The sequence chosen for splicing in vivo is that in the third
repeat. This sequence has high homology with the corresponding regions in
the other two repeats within intron 6, suggesting that sequence per se of the
intron-exon boundary is not the sole determinant of a functional splice
junction. This situation provides a natural anology to the observation that
when duplications are artificially introduced into a rabbit ~-globin gene the
effective acceptor site is the most 3' copy of the acceptor sequence
(Wieringa et al., 1983). Sequence divergence between Adhl and Adh2 is so
high in intron 6 that we cannot determine whether Adh2 once also had this
triplication. We consider it unlikely since the mutation rate which has
allowed divergence of the non-functional intron 6 sequences between Adhl
A. Adhl A A C C C C A G C A GAT T C GA.
V J . .: :
Adh2 AACCCAGCAAAATACGA.
B.
.... CTTGGTATAATCACT ..... 56 bp ... TTTTTTCAGCTTGGAAGTTTGGTTGCACTGGCA
CTTGGTCTAATAACT ...... 56 bp ...... TTTTTTCAGCTAGGAAGTTTGGTTGCACTGGC
CTTGGACTAATAACT .... 56 bp ..... TTTTTCAGCTAGGAAGTTCGGTTGCACTG .......
4 I
Intron 6 Exon 7
Fig. 4. A. Comparison of Adh 1 and Adh 2 sequence homologies in a portion of exon

5 where the genes differ by a likely double-frameshift (initial mutation followed by
compensating mutation). B. Intron 6/exon 7 junction of Adh 1 showing triplicate
nature. of base sequence
and Adh2 should also have obscured the triplication in Adhl, unless some
rectification mechanisms were operating.
Except for an 11 bp region which includes the TATA box and some
regions of 8 bp or less, the 5' and 3' regions of the Adhl and Adh2 genes
show little or no homology. The 5' non-translated region is 99 bp for Adhl
and 126 bp for Adh2. The 3' non-translated regions are approximately 300
to 400 bp long for both genes, depending on the allele under consideration.
Some transcription signals can be recognized in these regions (discussed
below) but in general it appears that while the overall structure and protein
coding sequences of Adhl and Adh2 are highly conserved, the non-trans-
lated sequences have almost completely diverged since the duplication
event occurred.
Structural information on gene organization is available for Adh genes
from two other species. Two genomic clones have been isolated from
barley by their cross hybridization with a maize Adhl probe (M. Trick and
K. Edwards, pers. comm.). Although neither of them contains a complete
gene, sufficient data is available to make some comparisons. One clone
obviously contains an Adh sequence, having high homology to both the
maize Adhl gene (80 % identity over 222 nucleotides) and maize Adh2
(79 %). It also has 78 % homology with another barley Adh gene. This
second barley clone contains an Adh2-like sequence with 89 % homology to
maize Adh2 but only 80 % homology to Adhl. In both barley genes the
introns which have been sequenced are located at identical positions to
those seen in the maize Adh genes.
An Arabidopsis Adh gene has also been fully sequenced (C. Chang and
E. Meyerowitz, pers. comm.). The coding sequence is 73 % and 72 % hom-
ologous with those of maize Adhl and Adh2 respectively. The predicted
amino-acid sequence is 81 % and 79 % conserved with the maize sequences.
Six of the nine introns found in maize are also present in the Arabidopsis
gene but differ in size and sequence, generally being smaller. Three introns
(4, 5 and 7) which are present in the maize genes are entirely absent from
the Arabidopsis gene. Presumably this reflects a mechanism by which Arab-
idopsis, one of the lowest nuclear C-value plants, is able to streamline its
genome.
v. Three-Dimensional Structure of ADH Enzymes

The amino acid sequences of the maize ADHI and ADH2 enzymes have
been derived from the nucleotide sequences of the genes (Dennis et ai.,
1984, 1985). They have 20 % amino-acid sequence homology with yeast
ADH and an even higher level of homology, 50 %, with horse liver ADH.
The three dimensional (3D) structure of the horse-liver enzyme has been
determined by X-ray diffraction and, considering its homology to the
maize enzymes, it can be used as a basis for 3D structuring of the maize
enzymes (Branden et ai., 1984). Computer modelling and physical model
building have enabled the complete 3D maize ADH structure to be pre-
dicted and the functional domains of the enzyme to be identified (Fig. 3,

Branden et al., 1984). The catalytic domain extends from residue 1 to 174
and 318 to 374 and the coenzyme binding domain roughly comprises res-
idues 175 to 315. The coenzyme binding domain is separated from the main
part of the catalytic domain by intron 4 and from the carboxy terminal
fragment of the catalytic domain by intron 9. The coenzyme binding
domain is further separated by intron 7 into two structurally similar (a~)3
domains. There is evidence that the two (a~)3 units are themselves struc-
tural modules; one of them is subdivided by introns 5 and 6 into three
similar supersecondary structural units, while the other is separated by
intron 8 into one a~ and one (a~)2 units. In the ancestor of the maize gene a
tenth intron may have been located between the now adjacent units in (a/3)2
region and its absence could be due to loss during evolution of the gene.
Certainly loss of introns can occur, as evidenced by the loss of three
introns during evolution of an Arabidopsis Adh gene (see above).
Branden et al. (1984) suggest that, as with most eukaryotic dehydroge-
nases, maize Adh may have evolved as two separate domains, the catalytic-
and coenzyme-binding domains, which later fused. The basic building
block of the coenzyme-binding site was a unit of about twenty amino acids.
Three such blocks were fused together to form a mononucleotide-binding
domain which later duplicated to form the basic structure of the NAD-
binding domain present in all characterized dehydrogenases.
From a consideration of the 3D structures of the proteins and a com-
parison of maize and horse-liver ADHs the following observations can be
made (H. Eklund and C.-I. Branden, pers. comm.):
1. The inter-subunit contact area is very different in the maize enzymes
to that in the horse-liver enzyme. However, it is conserved between maize
ADHI and ADH2. This is compatible with the fact that association can
occur to produce ADHlI ADH2 heterodimers.
2. Residues that are essential for coenzyme or catalytic activity can be
identified from an analysis of which amino acids are conserved between
maize ADHI and ADH2 and horse-liver enzyme. The position of these
regions in the molecule can be determined from the 3D model. Thus, ADH
provides a system in which fundamental studies on structure-function rela-
tionships of proteins can be made.
3. The different substrate specificities and inhibitor kinetics of the
maize and horse-liver enzymes can be accounted for on the basis of the 3D
structure. At position 48 there is a substitution of a threonine in the maize
ADHs for a serine in the horse, and this probably accounts for most of
these effects since it is located in the cleft of the catalytic domain close to
important functional groups.
4. The substitution of amino acids between the products of Sand F
alleles of maize Adhl (ala at 126 for gly, and asp at 360 for asn) has no
effect on the catalytic activity as both these residues are on the surface of
the catalytic domain. The ten fold difference in activity between ADHI
and ADH2 is probably explained by the substitution at position 228 of lys
byarg.
VI. Genetic Change Around and Within Adh Genes in Maize
One of the strengths of the Adhi and Adh2 genetic system of maize is the
number of mutants and variants available. These have been provided
largely by Drew Schwartz of Indiana University and his students and asso-
ciates. Knowledge of the Adhi gene structure together with the molecular
characterization of a number of these mutants has led to an understanding
of the structural nature of transposable elements, allelic variation, gene
duplication and somaclonal variation in maize.
A. Allelic Variation
Two naturally occurring variants of AdhI, Adhi-iS and Adhl-IF(Schwartz,
1966), encode, respectively, slow and fast anodally migrating electro-
phoretic allozymes of Adhi. The altered electrophoretic migration is due to
a single charge change (Kelly and Freeling, 1980) which we have deter-
mined from the derived amino-acid sequence of the gene sequence to be
due to a change from aspartate to asparagine (Sachs et al., 1985).
In addition to the difference in electrophoretic mobility, Adhi-iS and
Adhl-IF differ from each other in their tissue specific expression
(Schwartz, 1971; Woodman and Freeling, 1981). Expression of the
ADHI-IS polypeptide is higher than that of F in the scutellum of the
mature kernel. However, in both root and shoot tissue of anaerobically
induced seedlings, ADH-lF is found at approximately twice the level of
ADHl-lS. Another difference between the two alleles is that Adhi null
mutants derived from the IS allele recombine among themselves and so do
a series of null mutants derived from the IF allele. However, intragenic re-
combination between the Adhl-I Sand AdhI-I F alleles has not been
detected (Freeling, 1978). A further difference between the two alleles is
that, at least in the standard AdhI-F line (BKF), there are two different
length messenger RNAs but in the standard AdhI-S line (BKS) there is
only one (Gerlach et al., 1982).
Cloning and sequence comparison of these two alleles has shown them
to differ mainly at the 3' end. The coding regions and the introns are very
highly conserved as is the 5' nontranscribed region up to about 1200 bases
5' of the coding region. There are differences at the 3' end of the gene, even
in the transcribed region. This accounts for the two mRNAs in Adhi-iF
and one in IS. One of the main differences between the two alleles is the
presence of an insert with the form of a transposable element remnant
having terminal inverted repeats flanked by a short direct duplication of a
genome segment in AdhI-IFbut not in Adhl-IS. This extra DNA contri-
butes extra sites for poly A addition in the AdhI-I F transcript. The other
differences between the 3' ends of the IF and the IS allele are a tandem
duplication of 86 bp in IF and one of 14 bp in IS. There are also 108 bp
present in IS that are absent from IF and, at the 3' end of the gene, the
sequences of both alleles diverge completely.
From sequence analysis we know that the coding and immediate
upstream regions (1 kb) remain fairly constant, but that the 3' ends diverge
close to the coding region. Further from the conserved region, Southern
analysis suggests extensive sequence variation; restriction enzyme poly-
morphism is the rule when the two alleles are compared in these regions
(Johns et al., 1983; Sachs et al., 1986). We do not know whether these
flanking sequences are sufficiently different that a synaptonemal complex
cannot be formed and, therefore, that the alleles cannot recombine. Since
the main difference between the two alleles is at the 3' end, one suggestion
is that tissue specific expression may involve sequences in the 3' region.
The differences in the 5' region of the gene are minimal and we cannot
point to any sequence difference that appears likely to produce the dif-
ferent expression, although single base changes or small additions/ dele-
tions could be significant. Thus, from the analysis of the IS and IF alleles
we have a picture of the sequence differences between two naturally
occurring allelic variants in maize.
B. Somaclonal Variation
Plants which have been regenerated following a cycle of tissue culture
show variation (somaclonal variation) and such variation has generally
been described for morphological characters (Larkin and Scowcroft, 1981).
In order to investigate the molecular basis of somaclonal variation, muta-
tions arising in the Adhl gene of maize in tissue culture have been sought
and one has been isolated (Brettell et al., 1986). This mutant has an altered
electrophoretic mobility for the Adhl locus but retains full enzymatic
activity. Both Southern blot analysis and mapping of the mutant allele
show no obvious size difference between the gene and its progenitor. This
holds for all of the restriction fragments examined. It is likely that the
change in charge in the mutant is due to a point mutation in the coding
region. The difference between the mutant and the progenitor will be inves-
tigated at the nucleotide level to determine the mutational event. A mean-
ingful comparison can be made because we know the complete nucleotide
sequence of the progenitor Adhl gene.
C. Gene Duplication
There is a naturally occurring duplication of the Adhl-FC m locus
(Schwartz, 1973), two electrophoretic products being produced from the
same gene. The F polypeptide has full activity while crn has much lower
activity but is more heat stable. The coding regions specifying the two poly-
peptides cannot be separated by recombination, reflecting either the
closeness of the two alleles or else a block in recombination similar to the
block in intragenic recombination seen in the Adhl-F/Adhl-S comparison.
Mutation studies definitely show two coding regions since single mutations
affect either Crn or Fbut not both. Southern blot analysis of the FC rn allele
following BamHI digestion shows a single fragment of 9 kb different to
that seen from either the standard S (11 kb) or the standard F(7 kb) alleles.
When this 9 kb Bam fragment was cloned and mapped it was shown to
carry only a single copy of the Adhl gene. em also resides on a 9 kb Bam
fragment. Further Southern analysis showed that the duplicated region
extends for more than 14 kb. Thus, although Fern is genetically a dupli-
cation, the two copies of the structural gene must be separated by more
than 14 kb. This separate nature of the F and em portions of the dupli-
cation had been suggested already by the mutational analysis of Birchler
and Schwartz (1979). More extensive studies will be needed to show the
exact limits of the duplication and how it occurred.
D. Ds Element Mutations in Adhl

Knowing the sequence of the Adhllocus has enabled also the isolation and
sequencing of Ds controlling elements. A Ds controlling element mutation
of the Adhl gene (allele designation Adhl-Fm335) was isolated by
Osterman and Schwartz (1981). Comparison of the progenitor, mutant and
revertant alleles identified the Dsl element and led to the first sequencing
of a controlling element (Peacock et al., 1983; Sutton et al., 1984). In Fm335
the Dsl element was inserted into the 5' untranslated region of the gene
and caused about a 90 % decrease in ADHI activity. Analysis of the 405 bp
Ds element showed 11 bp inverted repeats at the termini and that it had
caused an 8 bp direct duplication of the genomic DNA at the site of
insertion. Between the 11 bp inverted repeats the internal sequence of the
Ds element is AT rich (80 % AT) and does not have any recognizable
coding capacity. S, mapping and Northern analysis suggest that much of
the Ds element behaves as an intron which is processed out of the tran-
script. It is possible that the unprocessed RNA transcript is rapidly
degraded since the mRNA level is much lower than normal (1 %) and there
is no obvious high level of precursor transcripts. Examination of a number
of revertants of this mutation showed that the Ds element does not excise
precisely and much or all of the 8 bp target site duplication remains,
although in altered form (Sachs et al., 1985; Dennis et al., in preparation).
During excision the bases bordering the site of insertion are either deleted
or undergo a complementary transversion. The study of a number of these
revertants provides the rules and models for the excision of Ac/Ds con-
trolling elements (Peacock et al., 1984; Saedler and Nevers, 1985).
Starting with the same original line used by Osterman and Schwartz
(1981), another Ds element mutation has been isolated at the Adhllocus
(Doring et al., 1984). Molecular characterization showed it to be a 1.5 kb
element derivative of the Ac9 family of Ac/Ds controlling elements which
has inserted into the coding region of the gene (see Doring and Starlinger,
1984).
E. Robertson's Mutator
Having an isolated and characterized Adhl gene has enabled characteri-
zation of Robertson's mutator elements. "Robertson's mutator" is a phe-
nomenon in which lines carrying the mutator activity give rise to mutants at
defined genetic loci at high frequency (Robertson, 1978). Mutations of
Adhl have been isolated and characterized (Strommer et al., 1982). One
such mutant allele, Adhl-S3034, was selected phenotypically as an allyl
alcohol resistant pollen grain from a stock containing Robertson's mutator.
S3034 has about 40 % normal ADHI activity and reverts to full activity at
about one hundred times the normal mutation rate. Mutant derivatives that
express 13 % and 0 % Adhl (S3034 A) protein were further selected. Com-
parison of the Adhl-l S progenitor and the S3034 mutant and its derivative
showed the presence of a 1.5 kb insert which was identified as a
Robertson's mutator element inserted into the first intron of the Adhl gene
(Bennetzen etal., 1984). The S3034 A derivative with zero Adhl activity
results from a deletion extending from the element into the first exon and
removing some of the coding region. The Robertson's mutator element has
been sequenced and is being used for transposon mutagenesis in attempts
to clone other genes.
VII. Approaches to the Mechanism of Adh Gene Regulation
As with any other plant or animal gene, there are three paths towards a
better understanding of Adh gene regulation. First, direct comparisons of
DNA sequences of different but functionally or physiologically related
genes give an indication of the gene regions conserved during evolution,
and hence of some possible functional significance. Second, correlations
between mutant phenotypes and the physical lesions found in natural or
induced mutations of genes can suggest a function for certain sequences.
Finally, this process can be refined by the specific in vitro mutation of
cloned genes and their re-insertion into the genome to test the effect of the
mutation on in vivo expression. Each of these paths complements the other
and will provide a picture of the molecular nature of the controls on the
expression of the Adh genes.
A. DNA Sequence Comparisons

Adhl and Adh2 share many biochemical and genetic characteristics that
suggest their origin from some ancestral gene duplication and subsequent
evolutionary divergence (Dennis et al., 1985). A direct comparison at the
sequence level should reveal those parts of the gene responsible for similar-
ities in expression and regulation, since they are more likely to be con-
served. As already indicated, the homologies within the coding regions of
Adhl and Adh2 are very high but elsewhere there is very little overall ho-
mology between the genes. The corresponding introns bear little
resemblance to one another except, as expected, around the intron-exon
junctions where specific splicing signals are to be found. It is in the non-
conserved 5' and 3' region that we should first look for sequence features
which might be responsible for transcription and translation rates as well
as for 3' polyadenylation signals. SI mapping and cDNA cloning have

indicated that there are mUltiple polyadenylation sites in AdhJ: four for
Adhl-1S and seven for the Adhl-1F allele (Sachs et al., 1985). In contrast,
Adh2 has a single polyadenylation site. Analysis of the sequence near the
polyadenylation sites in Adhl-1Fand Adhl-1S alleles reveals no conserved
sequence that could serve as a polyadenylation signal. The sequence
AAT AAT, on the other hand, is located 20 bp upstream of the single poly-
adenylation site of Adh2 mRNA. This is similar to the AATAAA signal
found 15-30 bp upstream of the polyadenylation site of animal genes
(Proudfoot and Brownlee, 1976).
In the regions 5' to the start of translation, several oligonucleotide ho-
mologies between Adhl and Adh2 have been identified by computer
analysis of the DNA sequences. The translation start point itself is within a
short region of homology (Fig. 5). This sequence, GGAATGGC, corre-
sponds to the conserved sequence for translation start sites in animal genes
(Kozak, 1981). Nearby, within the 5' untranslated region, there is a region
of homology located about 45 bp upstream of the translation start of each
gene. The role of this sequence is unknown as yet but it may serve as a
recognition signal to the translation machinery of the cell, since anaerobic
mRNAs are translated selectively and preferentially during anaerobiosis
despite the persistence of non-stress mRNA. Alternatively, it may be
important for Adh mRNA stability, since mutations in this region severely
affect mRNA levels and transcription rates (see below).
There is a cluster of three short stretches of homology associated with
the transcription start for the two Adh genes of maize which are probably
important for the precise location of the beginning of the mRNA (Fig. 5).
Transcription is initiated at an A within a pyrimidine rich tract, which
resembles the beginning of animal genes (Corden et al., 1980). The longest
stretch of homology outside of the coding region of the Adh genes occurs
about 35 bp upstream of the transcription start (Dennis et al., 1985). This
sequence,. . . ACCACTATATAAATCAG. . ., contains a classical TATA
Approx. -30 Transcription 5' Noncoding Coding
r:- C-
=-
1 . CTATATATA-C ... TTTTCTTCCTCAC ........................... GTTCTTGGAGTGG ............ GCAATG
I
2 . . . CTATATAAATC ..... TTTTC~~CTCACAN4cAT ..... TTCGTAACTGGTGA ... ~~iiTG
3. . .......... not known ................................... GTTCTTGGAGTGAAATG

4 . . . CTATATAAATC ..... TTTTCT~CACA ....... TTCTTCACTGTTGA ................... ~TG
~ I
1: Maize Adh2 (Dennis et al., 1985)

2: Maize Adh1 (Dennis et al., 1984)
3: Barley Adh3 (M.Trick and K.Edwards, pers.comm.)
4: Arabidopsis Adh (C.Chang and E.Meyerowitz, pers.comm.)
Fig. 5. Some of the sequence homologies detectable in comparisons of the 5' leader
and upstream sequences of Adh genes from maize, barley and Arabidopsis
box characteristic of animal viral and cellular genes (Breathnach and

Chambon, 1981). In animal systems at least, the TATA box is primarily
involved in the fixation of the site of transcription initiation to within the
narrow pyrimidine-rich region already described (Corden et ai., 1980;
Benoist and Chambon, 1981; Grosveld et ai., 1982). A TATA-like sequence
has been found in a similar position in all plant genes so far examined, and
its strong conservation, together with some flanking sequences, between
Adhl and Adh2 confirms its importance in plant cells for transcription by
RNA polymerase II.
In some animal genes other sequences upstream from the TATA box
have been also implicated in efficient transcription in vivo. In the globin
genes both the TATA box and the so-called CCAAT box at -80 are
essential for high levels of transcription. However, the animal gene con-
sensus CCAAT sequence can be rarely fitted with any confidence to
sequences in similar places in plant genes (Messing et ai., 1983), and
although we had previously assigned the sequences GGCCAAACC at -90
in the Adhl gene as a putative CCAAT sequence (Dennis et ai., 1984) there
is no homologous sequence in the Adh2 gene (Dennis et aI., 1985). Unless
such a sequence is involved in tissue-specific expression, which differs
between Adhl and Adh2, it should have been conserved like the TATA box
region. In genes where the CCAAT box is functionally significant, e. g. the
beta and alpha-globin genes, both it and the TATA box have been con-
served during evolution (Efstratiadis et ai., 1980).
Regions further upstream in Adhl and Adh2 have been scanned in an
attempt to locate any homologous sequence(s) that may be responsible for
the anaerobic induction of these genes. Genes such as the metallothionens
(Mayo et ai., 1982) and the heat-shock genes (Pelham, 1982), which are
coordinately induced in response to a given stimulus, have been found to
include a common regulatory sequence 5' to the coding sequence. When
placed upstream of any other gene, these sequences confer on that gene the
inducible response of the gene from which it was derived. Comparisons of
quite extensive regions of the 5' ends of Adhl and Adh2 revealed very little
sequence homology. A short eight base sequence, CACCTCCC, at -170 in
Adh2 and -200 in Adhl is a candidate for an anaerobic control signal
(Dennis et ai., 1984 b). If the homology criteria are relaxed, two more
extensive regions of imperfect homology, each extending about thirty
bases, can be seen at about -80 and -160 in both genes.
Although these sequence comparisons reveal regions of homology
between the maize Adhl and Adh2 genes, their importance in the regulation
of Adh gene expression can only be determined by functional assays. Some
indication of importance may come from comparisons with Adh genes of
other plant species as they are characterized, but it is important to note that
different nucleotide signals may have evolved to perform the same
function in other species.
Some sequence data of the 5' flanking region has been obtained for the
barley Adh2-like gene (M. Trick, unpublished) and an Arabidopsis Adh gene
(c. Chang, pers. comm.). For the barley gene, the sequence information
includes the translation start and some 80 bp of untranslated region. There

is a 12 bp perfect match to the conserved sequence seen in the 5' leader of
Adhl and Adh2 of maize although the location is somewhat different
(Fig. 5). S 1 mapping of this barley gene has defined the start of tran-
scription at 70 bp upstream from the start of translation at the sequence
TCATCAGCAA which shows considerable homology to the Adh2 tran-
scription start (TCCTCACCAA), although the untranslated region is much
shorter (126 bp for Adh2).
For the Arabidopsis gene, the homology again involves the conserved
sequence found in the 5' non-translated sequence of the maize genes and,
like barley, this sequence is found very close to the translation start point
(Fig. 5). A sequence ACATCACAA, which has some homology with the
maize and barley transcription start regions, is found approximately 50 bp
upstream of the translation start. However, it is not yet known where tran-
scription begins in the Arabidopsis Adh gene.
B. Mutations in Adh Which Affect Expression

Preliminary results show that the control of induction of the Adhl gene is
at the level of transcription. This has been done by isolating nuclei and
measuring "runoff' transcription (L. Beach, pers. comm.). Under an-
aerobic conditions there is an increased amount of incorporation of radio-
active precursors into Adh-RNA. This is reflected in a 50-fold increased
steady-state level of Adhl mRNA. Adh2 mRNA steady-state levels also
increase 50-fold upon anaerobiosis but we have not done the experiments
to discover whether this is a transcriptional response.
Some of the mutants of Adhl have both a decreased mRNA level and a
decreased transcription rate. For example, in the Ds mutant allele
Adhl-Fm335 there is a very low steady-state level of Adhl specific mRNA.
The Dsl element has inserted into the 5' untranslated region of the gene,
and nuclease S 1 mapping shows that transcription still starts at the same
point as Adhl. These results show that insertion of 402 bp about 45 bp
upstream from the translation start in the 5' untranslated region does not
affect the point of transcription initiation, so the 402 bp is not providing a
new promoter. Although there is a marked effect on the amount of mRNA,
preliminary results from "runoff' transcription studies (L. Beach, unpub-
lished) show that the transcription rate is maintained. Therefore, it is not
the amount of RNA made but its stability that is affected. S 1 nuclease data
and sequence analysis suggest that the Ds element acts in the Adhl gene as
an extra intron which may not be excised at the normal rate, thus possibly
affecting the stability of the RNA. In revertants of the Adhl-Fm335 allele
which have full activity we found that the extra 6-8 bp present in the same
position in the 5' untranslated region (the remnant of the duplication
formed upon insertion) and present in the mRNA have no effect on either
the level of RNA or of ADH enzyme.
A null derivative of the Adhl-Fm335 allele has been isolated with a
deletion of 70 bp from the insertion of the Ds element at position 45 into
the first exon of the gene. As expected, this mutant produces no enzyme
because it is missing a portion of the coding region including the start of
translation, but, more important, there is also very little mRNA. Again this
could be due to instability of the RNA or a decreased rate of transcription.
The notable feature of these two mutations, the Adhl-Fm335 Ds insertion
and the null derivative with a deletion, is that the 5' non-coding region
affected is close to a sequence conserved between Adhl and Adh2.
In another case, the insertion of a Robertson's mutator element into the
first intron of Adhl (Strommer et aI., 1982) lowers the level of mRNA but
the cause of this decrease is not known.
Overall, we see that some insertions into the 5' untranslated region or
the introns of the maize Adhl gene do have an effect on expression,
altering the rate of transcription or the stability of the message, while
certain smaller insertions have no effect. Many more mutants and variants
of the Adhl locus are still to be examined and molecular analysis of these
should assist in our understanding of gene regulation.
For Adh2, the only mutant which has been analysed is the Adh2-33
allele found as a naturally occurring null mutation in the line Knobless
Wilbur Flint (Dlouhy, 1980). This gene has a 200 bp insertion in the 3'
region downstream of the polyadenylation site (J. Ellis, unpublished) and
this may be the reason for the lack of detectable message for this mutation
(Dennis et ai., 1985). However, recent results have shown that this 3'
insertion does not affect polyadenylation activity in tobacco when tested in
T-DNA transformation experiments (S. Ben-Tahar, pers. comm.), so there
may be some other reason for the null phenotype of the Adh2-33 allele.
A transformation system is needed for studying expression of natural
and in vitro mutated gene segments of Adhl genes and for defining those
elements important for regulation and expression.
VIII. In Vivo Expression of Adh
Sequence comparisons and mutant studies have given some insights into
the possible nature and location of regulatory signals within and adjacent
to the Adhl structural gene, but regulatory functions for these sequences
must, at best, be considered tentative. Confirmation of an in vivo function
for any specific sequence will only come from in vivo expression studies on
genetically-engineered promoters or genes. Such analyses using rearrange-
ments of DNA segments adjacent to genes in animal cells have identified
signal sequences responsible for regulating expression, e. g. during heavy
metal induction (Mayo et ai., 1982) and heat shock induction (Pelham,
1982) of specific genes or sets of genes. However, in plants it has been dif-
ficult to put in vitro modified genes back into cells to assess the effects of
the alterations on gene expression. The difficulties are now being resolved
and allow the study of Adh gene regulation on three broad fronts: first, and
possibly most important, in a completely homologous system involving
maize protoplasts, pollen or endosperm as recipients for modified Adh
genes; second, in an heterologous system with, for example, tobacco, and

finally, the introduction of Adh genes into animal cells.
A. Attempts at Expression in Cereal Tissue Culture Cells

It is now possible to prepare protoplasts from callus or suspension cultures
of a limited number of cell lines of maize, wheat and rice which retain the
potential for cell wall regeneration and continued cell division to reform a
callus but will not, unfortunately, regenerate into plants. Transformation of
cereal protoplasts may lead to either short-term (transient) expression or
stable integration and expression of introduced genes. Both require the
uptake of vector DNAs into protoplasts but only the latter demands the
survival and growth of protoplasts for extended periods of time under a
specific selection regime. The transient expression assays will be useful in
the optimization of vector entry into protoplasts (since many variations can
be tested in a short time) and in the study of genes expressed in proto-
plasts. Stable transformation will be essential for any studies of gene regu-
lation and genetic engineering of plants.
A transient expression system is now being used to examine Adhl
expression. A chimeric gene has been constructed with a maize Adhl
promoter fragment, extending from -1 (relative to the start of translation)
to approximately 1.5 kb upstream, fused with a bacterial chloramphenicol
resistance (CAT) gene. We are using protoplasts of the maize line of
Chourey and Zurawski (1981) in which we have already shown transient
expression of the CAT gene with the Ti plasmid nopaline synthase
promoter. As yet we have no data on the expression of this Adh construct.
Stable transformation has not as yet been achieved with maize, but has
been reported recently for Triticum monococcum (Lorz et al., 1985). Wheat
protoplasts were treated with polyethylene glycol and a vector containing a
chimeric gene (pNOS-Kan) known to express a bacterial aminoglycoside
phosphotransferase (NPT-II) activity when introduced into dieot plants
(Herrera-Estrella et al., 1984). In tobacco this gene confers resistance to as
much as 500 J.lg/ml kanamycin, a dose twenty times greater than the
minimum lethal concentration. Experiments are under way to establish an
effective selection system in maize using kanamycin and G418 (K. Danna,
unpublished), attempting to select for resistant cells after treatment of pro-
top lasts with DEAE-Dextran and vector DNAs containing various combi-
nations of pNOS-Kan, pAdh-Kan, with Ac and Ds controlling element
sequences.
The most interesting properties of Adhl and Adh2 (and most other
genes) are their spatial and temporal regulation in the whole plant. There is
thus considerable interest in the development of a transformation system in
which tissue-specific and developmental aspects of Adh expression can be
studied at the molecular level. There has been some suggestion of the
transfer of Tripsacum genetic characters into maize after the fertilization of
maize ovules with pollen treated with isolated Tripsacum genomic DNA
(D. Jewell, pers. comm.) but no molecular proof of actual gene transfer is
available. The same approach can be used with sequences such as the
cloned Adh genes as donor DNAs and defined mutants at the Adh loci as
recipients. Pollen from an AdhI- Adh2- plant treated with vectors con-
taining either the Adhl-F of Adhl-S genomic clones was used to pollinate
another AdhI- Adh2- plant or cultured ovules (J. Waldron, unpublished).
Three out of more than 1000 seeds set by this procedure and screened for
ADH activity on starch gels showed a low level of the appropriate Adh
allele. Unfortunately all three lacked vigour and died shortly after germi-
nation making confirmation of transformation impossible.
B. Testing Maize Adh Gene Activity in Dicotyledonous Plant Cells

The transformation of dicot plants provides a different perspective by
studying maize Adh gene regulation in a phylogenetically unrelated plant
species. Gene transfer based on the Ti-plasmid of Agrobacterium tumefa-
ciens is the most established technique for dicots and we have used this to
introduce maize Adh constructs into tobacco species.
Initially, the entire Adhl-S genomic clone, including considerable 5'
and 3' flanking sequences, was inserted into two different types of Ti-
plasmid and the resultant bacterial strains used to produce tumours on
Nicotiana tabacum and N. plumbaginifolia. Sterile tumour tissues contained
the intact maize gene integrated into tobacco DNA but the gene was not
expressed at the mRNA level (Llewellyn et al., 1985). Lack of expression
may arise from a defect at any step in the transcription or processing of
RNA, so other chimeric constructs, including the pAdh-CAT gene
(described above) and new constructs with slightly shorter maize promoter
fragments coupled to the octopine synthase gene (pAdh-OCS), have been
used to try to define the problem. In both cases 3' signals were those of the
nopaline synthase gene. No CAT or OCS activity was detected in tumours
or regenerated plants containing these chimeric genes (J. Ellis, unpub-
lished). The 3' end of Adhl-S tested in tobacco by substitution for the NOS
3' end in pNOS-CAT functioned as well or better than the NOS 3' end
(S. Ben-Tahar, pers. comm.). Polyadenylation occurs at the same polyaden-
ylation site as in the Adh2 gene in maize.
Although we cannot discount the possibility that a low level of RNA is
transcribed using the maize promoter but is rapidly turned over, the data
strongly suggest that the Adhl promoter is not functional in tobacco. An
identical problem has been encountered with other maize and wheat genes
introduced into tobacco and Petunia and this may be a more general phe-
nomenon.
C. Expression of Maize Adh in Animal Cells

The Adhl gene has been intoduced into monkey cells using SV40-based
vectors but was not transcribed at a detectable level (E. S. Dennis, unpub-
lished). A chimeric gene containing an Adhl promoter attached to a
kanamycin resistance gene was expressed, and at increased levels in the
presence of a viral enhancer sequence. Transcription was shown to start in

the same place as in maize but was not anaerobically inducible. The 3' end
of Adhl tested in monkey cells in a chimeric SV40 early gene promoter-
maize Adhl-F cDNA gene was functional but used a polyadenylation site
upstream from the true site. This new site was just beyond a sequence that
mimicked the animal AATAAA polyadenylation signal, but which is not
used in maize. Clearly, plant genes do not function in the same way when
taken out of their normal environment.
IX. Conclusions
It is obvious that the wealth of genetic information on the alcohol dehy-

drogenases of maize led to the cloning of their genes. From there we have
been able to structurally characterize the genes and use this information to
address other genetic phenomena such as the nature of allelic variation,
transposable elements and some indication of the basis of mutations in
plants. Although there is still much to be done to determine the structural
change in many of the defined Adh mutants and variants, the next major
area which must be developed is an understanding of the regulation of
these genes. This may be approached in two ways. First, by analysis of the
structure of natural and induced variants. Obviously this will provide clues
only by relating structural alterations to phenotypic changes. Second, and
more powerful, the suspected regulatory regions must be tested for activity
in vivo by preparing constructs for transformation. The constructs are being
prepared, the transformation system is required.
Acknowledgements
We wish to thank the many colleagues cited in the chapter who provided us
with information on their unpublished data. During preparation of this
manuscript W. L. G. was the recipient of a Harkness Fellowship at the Uni-
versity of California, Davis.
X. References
Bennetzen, 1. L., Swanson, 1., Taylor, W. C., Freeling, M., 1984: DNA insertion in
the first intron of maize Adhl affects message levels: cloning of progenitor and
mutant Adhl alleles. Proc. Nat. Acad. Sci., U.S.A. 13,4125-4129.
Benoist, C., Chambon, P., 1981: In vivo sequence requirements of the SV40 early
promoter region. Nature 290, 304-310.
Birchler, 1. A., Schwartz, D., 1979: Mutational study of the alcohol dehy-
drogenase-I FC rn duplication in maize. Biochem. Genet. 17, l173-1180.
Branden, C.-I. Eklund, H., Camillau, C., Pryor, A. 1., 1984: Correlation of exons
with structural domains in alcohol dehydrogentse. EMBO 1. 3, 1307 -131 O.
Breathnach, R., Chambon, P., 1981: Organisation and expression of eukaryotic

split genes coding for proteins. Ann. Rev. Biochem. 50, 349-383.
Brettell, R. I. S., Dennis, E. S., Scowcroft, W. R., Peacock, W. J., 1986: Molecular
analysis of a somaclonal mutant of maize alcohol dehydrogenase. Mol. Gen.
Genet. 202, 235-239.
Chourey, P. S., Zurawski, D. B., 1981: Callus formation from protoplasts of a maize
cell culture. Theor. Appl. Genet. 59, 341-344.
Cooper, P., Ho, T. H. D., 1981: Heat shock proteins in maize. Plant Physiol. 71,
215-222.
Corden, J., Wasylyk, B., Buchwalder, A., Sassone-Corsi, P., Kedinger, C.,
Chambon, P., 1980: Promoter sequences of eukaryotic protein coding genes.
Science 209, 1406-1414.
Dennis, E. S., Gerlach, W. L., Pryor, A. J., Bennetzen, J. L., Inglis, A., Llewellyn,
D., Sachs, M. M., Ferl, R. J., Peacock, W. J., 1984: Molecular analysis of the
alcohol dehydrogenase (AdhI) gene of maize. Nuc. Acid Res. 12,3983-4000.
Dennis, E. S., Sachs, M. M., Gerlach, W. L., Finnegan, E. J., Peacock, W. J., 1985:
Molecular analysis of the alcohol dehydrogenase-2 (AdhZ) gene of maize. Nuc.
Acid Res. 13, 727-743.
Dlouhy, S. R., 1980: Genetic, biochemical and physiological analysis involving the
Adh210cus of Zea mays. Ph. D. Dissertation. Indiana University.
Doring, H.-P., Starlinger, P., 1984: McClintock's controlling elements now at the
DNA level. Cell 39, 253-259.
Doring, H. P., Freeling, M., Hake, S., Johns, M. A., Kunze, R., Merckelbach, A.,
Salamini, F., Starlinger, P., 1984: A Ds- mutation of the AdhI gene in Zea mays.
Mol. Gen. Genet. 193, 199-204.
Efstradiatis, A., Posakony, J. W., Maniatis, T., Lawn, R. M., O'Connell, c., Spritz,
R. A., De Riel, J. K., Forget, B. G., Weissman, S. M., Slightom, J. L., Blechl, A.
E., Smithies, 0., Barelle, F. E., Shoulders, C. C., Proudfoot, N. J., 1980: The
structure and evolution of the human B-globin gene family. Cell 21, 653-668.
Ferl, R. J., Dlouhy, S. R., Schwartz, D., 1979: Analysis of maize alcohol dehydro-
genase by native SDS two-dimensional electrophoresis and autoradiography.
Mol. Gen. Genet. 169, 7 -12.
Ferl, R. J., Brennan, M. D., Schwartz, D., 1980: In vitro translation of maize alcohol
dehydrogenase (EC 1.1.1.1.): Evidence for the anaerobic induction of mes-
senger RNA. Biochem. Genet. 18,681-692.
Freeling, M., 1973: Simultaneous induction by anaerobiosis or 2,4-D of multiple
enzymes specified by two unlinked genes: Differential ADHI-ADH2 expression
in maize. Mol. Gen. Genet. 127,215-227.
Freeling, M., 1974: Dimerization of mUltiple maize ADH's studied in vivo and in
vitro. Biochem. Genet. 12,407-417.
Freeling, M., 1976: Intragenic recombination in maize: Pollen analysis methods
and the effect of parental AdhI+ isoalleles. Genetics 83, 701-717.
Freeling, M., 1978: Allelic variation at the level of intragenic recombination. Ge-
netics 89, 211-224.
Freeling, M., Bennett, D. C., 1986: Maize Adhl. Annual Review of Genetics. In
press.
Freeling, M., Birchler, J. A., 1981: Mutants and variants of the alcohol dehy-
drogenase-I gene in maize In: Setlow J. K., Hollaender A., (eds.), Genetic
Engineering: Principles and Methods, pp. 223-264. New York: Plenum.
Freeling, M., Cheng, D. S.-K. 1978: Radiation-induced alcohol dehydrogenase
mutants in maize following allyl alcohol selection of pollen. Genet. Res. 31,
107-130.
Freeling, M., Schwartz, D., 1973: Genetic relationships between the multiple
alcohol dehydrogenases of maize. Biochem. Genet. 8, 27 -36.
Gerlach, W. L., Pryor, A. J., Dennis, E. S., Ferl, R. J., Sachs, M. M., Peacock, W. J.,
1982: cDNA cloning and induction of the alcohol dehydrogenase gene (Adhl)
of maize. Proc. Nat. Acad. Sci., U.S.A. 79, 2981-2985.
Grosveld, G. C., de Boer, E., Shewnamer, C. K., Flavell, R. A., 1982: DNA
sequences necessary for transcription of the rabbit p-globin gene in vivo.
Nature 295, 120-126.
Hageman, R. H., Flesher, D., 1960: The effect of anaerobic environment on the
activity of alcohol dehydrogenase and other enzymes of com seedlings. Arch.
Biochem. Biophys. 87, 203-209.
Hake, S., Taylor, W. C., Freeling, M., 1984: Molecular analyses of genetically stable
mutants of the Adhl gene of maize. Mol. Gen. Genet. 194, 42-48.
Herrera-Estrella, L., Van den Broeck, G., Maenhaut, R., Van Montagu, M., Schell,
J., 1984: Light-inducible and chloroplast-associated expression of a chimaeric
gene introduced into Nicotiana tabacum using a Ti plasmid vector. Nature 310,
115-120.
Johns, M. A., Strommer, J. N., Freeling, M., 1983: Exceptionally high levels of re-
striction site polymorphism in DNA near the maize Adhl gene. Genetics 105,
733-743.
Kelley, P. M., Freeling, M., 1982: A preliminary comparison of maize anaerobic
and heat-shock proteins. In: Schlesinger, M. J., Ashburner, M., Tissieres, A.
(eds), Heat Shock From Bacteria to Man. pp. 315-329. Cold Spring Harbor,
New York: Cold Spring Harbor Lab.
Kelley, P. M., Freeling, M., 1984a: Anaerobic expression of maize glucose phos-
phate Isomerase I. J. BioI. Chern. 259, 673-677.
Kelley, P. M., Freeling, M., 1984b: Anaerobic expression of maize fructose-
1,6-diphosphate aldolase. J. BioI. Chern. 259, 14180-14183.
Kelly, J., Freeling, M., 1980: Purification of maize alcohol dehydrogenase-l allo-
zymes and comparison of their tryptic peptides. Biochim. Biophys. Acta 624,
102-110.
Kozak, M., 1981: Possible role of flanking nucleotides in recognition of the AUG
initiator codon by eukaryotic ribosomes. Nucleic Acids Res. 9, 5233-5252.
Larkin, P. J., Scowcroft, W. R., 1981: Somaclonal variation - a novel source of
variability from cell cultures for plant improvement. Theor. Appl. Genet. 60,
197-214.
Laszlo, A., 1981: Maize pyruvate decarboxylase: An inducible enzyme. Ph. D. dis-
sertation. University of California, Berkeley.
Lin, C. Y., Key, J. L., 1967: Dissociation and reassembly of polyribosomes in
relation to protein synthesis in the soybean root. J. Mol. BioI. 26, 237-247.
Llewellyn, D., Dennis, E. S., Peacock; W. J., 1985: The alcohol dehydrogenase
genes of maize: expression studies in tobacco cells. In: Van Vloten-Doting, L.,
Groot G. S. P., and Hall, T. C. (eds.), Molecular form and function of the plant
genome. pp. 593-607. New York: Plenum.
Lorz, H., Baker, B., Schell, J., 1985: Gene transfer tocereal cells mediated by proto-
plast transformation. Mol. Gen. Genet. 199, 178-182.
Mayo, K. E., Warren, R., Palmiter, R. D., 1982: The mouse metallothionen-l gene
is transcriptionally regulated by cadmium following transfection into human or
mouse cells. Cell 29, 99-108.
Messing, J., Geraughty, D., Heidecker, G., Hu, N. T., Kridl, J., Rubenstein, I., 1983:
Plant gene structure. In: Kosuge, T., Meredith, C. P., Hollaender, A. (eds.),
Gentic Engineering of Plants, pp. 211-277. New York: Plenum.
Okimoto, R., Sachs, M. M., Porter, E. K., Freeling, M., 1980: Patterns of poly-
peptide synthesis in various maize organs under anaerobiosis. Planta 150,
89-94.
Osterman, J. C., Schwartz, D., 1981: Analysis of a controlling element mutation at
the Adh locus of maize. Genetics 99, 267-273.
Peacock, W. J., Dennis, E. S., Gerlach, W. L., Llewellyn, D., Lorz, H., Pryor, A. J.,
Sachs, M. M., Schwartz, D., Sutton, W. D., 1983: Gene transfer in maize: Con-
trolling elements and the alcohol dehydrogenase genes. In: Downey, K. (ed.),
Proc. Miami Winter Symposium, pp. 311-325. New York: Academic Press.
Peacock, W. J., Dennis, E. S., Gerlach, W. L., Sachs, M. M., Schwartz, D., 1984:
Insertion and excision of Ds controlling elements in maize. Cold Spr. Harb.
Symp. Quant. BioI. 49, 347 -354.
Pelham, H. R. B., 1982: A regulatory upstream promoter element in the Drosophila
Hsp70 heat shock gene. Cell 30, 517 -528.
Proudfoot, N. J., Brownlee, G. G., 1976: 3' non-coding region sequences in eukary-
otic messenger RNA. Nature 263, 211-214.
Robertson, D. S., 1978: Characterization of a mutator system in maize. Mutat. Res.
51,21-28.
Sachs, M. M., Freeling, M., 1978: Selective synthesis of alcohol dehydrogenase
during anaerobic treatment of maize. Mol. Gen. Genet. 161, 111-115.
Sachs, M. M., Freeling, M., Okimoto, R., 1980: The anaerobic proteins of maize.
Cell 20, 761-768.
Sachs, M. M., Dennis, E. S., Gerlach, W. L., Peacock, W. J., 1986: Two alleles of
maize Alcohol dehydrogenase 1 exhibit 3' structural and Poly(A) addition Poly-
morphisms. Genetics, in press.
Saedler, H., Nevers, P., 1985: Transposition in plants: A molecular model. EMBO
J. 4, 585-590.
Scandalios, J. G., 1966: Genetic variation of Adh in maize. Genetics 54, 359.
Schwartz, D., 1966: The genetic control of alcohol dehydrogenase in maize; gene
duplication and repression. Proc. Nat. Acad. Sci., U.S.A. 56, 1431-1436.
Schwartz, D., 1969 a: An example of gene fixation resulting from selective
advantage in suboptimal conditions. Amer. Nat. 103, 479-481.
Schwartz, D., 1969b: Alcohol dehydrogenase in maize: Genetic basis for multiple
isozymes. Science 164, 585-586.
Schwartz, D., 1971: Genetic control of alcohol dehydrogenase - a competition
model for regulation of gene action, Genetics 67,411-425.
Schwartz, D., 1973: The Adhl-FCm "operon". Maize Genet. Coop. News Lett. 47,
53-54.
Schwartz, D., 1978: Allele-specific modification of the expression of the Adh2 gene
in maize by ethylene. Genetics 90, 323-330.
Schwartz, D., Endo, T., 1966: Alcohol dehydrogenase polymorphisms in maize-
simple and compound loci. Genetics 53, 709-715.
Schwartz, D., Laughner, W. J., 1969: A molecular basis for heterosis. Science 166,
626-627.
Schwartz, D., Osterman, J., 1976: A pollen selection system for alcohol dehy-
drogenase-negative mutants in plants. Genetics 83, 63-65.
Strommer, J., Hake, S., Bennetzen, J., Taylor, W. c., Freeling M., 1982: Regulatory
mutants of the maize Adh 1 gene caused by DNA insertons .. N ature 300, 542- 544.
Stuber, C. W., Goodman, M. M., 1983: Allozyme genotypes of popular and histori-
cally important inbred lines of corn, Zea mays L. ARS-ARR Southern Series 16,
1-29.
Sutton, W. D., Gerlach, W. L., Schwartz, D., Peacock, W. J., 1984: Molecular
analysis of Ds controlling element mutations at the Adhl locus of maize.
Science 223, 1265-1268.
Wieringa, B., Meyer, F., Reiser, J., Weissmann, C., 1983: Unusual splice sites
revealed by mutagenic inactivation of an authentic splice site of the rabbit ~
globin gene. Nature 301, 38-43.
Wignarajah, K., Greenway, H., 1976: Effect of anaerobiosis on activities of alcohol
dehydrogenase and pyruvate decarboxylase in roots in Zea mays. New Phytol
77,575.
Woodman, J. C., Freeling, M., 1981: Identification of a genetic element that con-
trols the organ-specific expression of Adhl in maize. Genetics 98, 354-378.
Chapter 5
The Molecular Genetics of Higher Plant Nitrate Assimilation
John L. Wray
Department of Biochemistry and Microbiology, University of St. Andrews, Irvine

Building, North Street, St. Andrews, Fife, KY 16 9 AL, Scotland
With 2 Figures
Contents
I. Introduction
II. The Nitrate Assimilation Pathway
A. Nitrate Uptake
B. Nitrate Reductase
C. Nitrite Reductase
III. Genetics of Nitrate Assimilation
A. Introduction
B. Nitrate Uptake Mutations
C. Nitrate Reductase Mutations
D. Nitrite Reductase Mutations
E. Regulatory Alterations in Nitrate Assimilation
F. Conclusions
IV. Applied Aspects
A. Somatic Hybridisation
B. Cloning Nitrate Assimilation Genes
C. Plant Gene Transfer Systems
D. Whole Plant Studies
V. References
I. Introduction
The nitrate assimilation pathway is the major point of entry of inorganic

nitrogen into organic combination in most crop plants growing in well-
aerated soils. This pathway converts nitrate to ammonium via a nitrate
uptake mechanism and two enzymes, nitrate reductase (NR) and nitrite
reductase (NiR), thereby making nitrogen available for a variety of biosyn-
theses, of which the most important quantitatively is protein synthesis.
102 J. L. Wray
Nitrate and hormone availability, light and end-products have all been
shown to playa role in the regulation of this pathway and they presumably
act together to match the in vivo rate of nitrate reduction to the physio-
logical status of the plant (reviewed in Beevers and Hageman, 1980; Sri-
vastava, 1980; Guerrero et al., 1981; Huffaker, 1982; Wallace and Oaks,
1984). However, these physiological and biochemical studies have told us
very little about the molecular mechanisms underlying the development
and regulation of this important pathway. Such information is critical if we
are to understand how the new techniques of molecular biology might be
applied to improve the efficiency of the nitrate assimilation pathway.
Over the last eight years or so there has been considerable progress in
the understanding of the genetics of nitrate assimilation, making it the best
characterised metabolic pathway in higher plants. These genetic studies are
an important first step in developing a molecular understanding of the
pathway and are reviewed here in some detail against a background of the
current biochemical information. I also consider some applications of
these studies. In some places, where appropriate, I have drawn on the vast
body of information available on the genetics of Aspergillus and Neuro-
spora nitrate assimilation. The genetics of higher plant nitrate assimilation
has recently been reviewed also by Dunn-Coleman et al. (1984) and
Kleinhofs et al. (1985).
II. The Nitrate Assimilation Pathway
A. Nitrate Uptake
Little is known about the molecular basis of nitrate uptake. It is clear that
uptake is an active process dependent on metabolic energy, since cyanide
and anaerobic conditions, as well as uncouplers of oxidative phoshory-
lation, are inhibitory (Rao and Rains, 1976). Nitrate stimulation of nitrate
uptake, after a lag in plants not previously exposed to nitrate (Minotti et
al., 1968 ; Jackson et al., 1973; Rao and Rains, 1976), and the inhibitory
effects of protein and RNA synthesis inhibitors, suggest that nitrate
induces a specific nitrate uptake system. Nitrate uptake rates follow a dual-
phase relationship with the concentration of nitrate available to the plant.
Both phases appear to be hyperbolic in accordance with Michaelis-Menten
kinetics and suggest the existence of at least two uptake mechanisms
operating at high and low nitrate concentrations respectively (Rao and
Rains, 1976; Doddema and Telkamp, 1979). Other studies emphasize the
role of nitrate efflux in determining net uptake rate (Jackson et al., 1976;
Clement et al., 1978; Deane-Drummond and Glass, 1983; Deane-
Drummond, 1982).
Net nitrate uptake is dependent on an influx term determined primarily
by external nitrate concentration and independent of internal nitrate con-
centration, and a variable efflux term directly proportional to internal
nitrate concentration (Deane-Drummond and Glass, 1983). Glass et al.
Genetics of Nitrate Assimilation 103
(1985) have argued that the reduction in net nitrate uptake associated with
nitrate loading of tissue is due to nitrate inhibition of tonoplast nitrate flux
into the vacuole. This leads to increased cytoplasmic nitrate concentration
and/or increased transport of nitrate to the shoot, together with increased
efflux of nitrate. There is, however, evidence to suggest that nitrate efflux is
not a futile "leak" of nitrate which is surplus to the requirements of the cell
but that a specific carrier is involved (Deane-Drummond, 1985). These
results have recently been interpreted in terms of a substrate cycle (Deane-
Drummond, 1986). How they relate to the earlier data showing the exis-
tence of dual phase net-uptake systems is unclear at present.
B. Nitrate Reductase
Higher plant nitrate reductases are flavohaemomolybdoproteins which ca-
talyse the two electron reduction of nitrate to nitrite (Hewitt and Notton,
1980). Almost all higher plants so far examined possess NADH-NR (EC
1.6.6.1) which has a pH optimum around 7.4 and a Michaelis constant for
nitrate and NADH of ca. 200 J.1M and 2 J.1M respectively (Beevers and
Hageman, 1980; Guerrero et al., 1981). Some plants also possess a bispe-
cific NAD(P)H-NR (EC 1.6.6.2) which uses either NADH or NADPH as
electron donor.
NAD(P)H-NR has been found in rice seedlings (Shen et al., 1976),
maize scutella (Campbell, 1978), maize roots (Redinbaugh and Campbell,
1981), soybean cotyledons (Orihuel-Iranzo and Campbell, 1980) and
soybean leaves (Evans and Nason, 1953; Jolly etal., 1976; Campbell, 1976;
Nelson et al., 1984). Biochemical and physiological evidence suggests that
the NAD(P)H-NRs are distinct species from the NADH-NRs with a higher
Km for nitrate (4 mM) and a lower pH optimum (6.5) (Campbell, 1976).
The two forms can be separated by chromatography on blue dextran
Sepharose (Redinbaugh and Campbell, 1981) and show different develop-
mental (Orihuel-Iranzo and Campbell, 1980) and induction (Shen et al.,
1976) patterns. The tropical legume, Erythrina senegalensis, is unique in
prossessing only NAD(P)H-NR (Stewart and Orebamjo, 1979). The
NAD(P)H-NRs may be smaller than the NADH-linked enzymes (Redin-
baugh and Campbell, 1981; Streit et al., 1985). Most work has been carried
out on the NADH-nitrate reductases and this is discussed below.
i) Molecular Weight and Prosthetic Groups
Attempts to determine the molecular weight of the holoenzyme subunits
have been complicated by their extreme sensitivity to proteolytic modifi-
cation (Brown et al., 1981; Wray and Kirk, 1981; Campbell and Wray,
1983; Wallace and Oaks, 1984). Their stability appears to differ between
species and even between genotypes of the same species. Purification in the
presence of proteinase inhibitors either reduces or eliminates molecular
weight species of 20,000-75,000 seen after SDS polyacrylamide gel elec-
trophoresis of several apparently homogeneous enzyme preparations
(Notton and Hewitt, 1979; Mendel and Muller, 1980; Campbell and Wray,
104 1. L. Wray
1983; Redinbaugh and Campbell, 1983) and indicates subunit values of

100,000 in barley cv. Golden Promise (Campbell and Wray, 1983), 100,000
(Kuo et al., 1980) or 110,000 (Kuo et al., 1982) in barley cv. Steptoe, 105,000
and 114,000 (Nakagawa et al., 1985) or 110,000 and 120,000 (Fido and
Notton, 1984) in spinach and 115,000 in squash (Redinbaugh and
Campbell, 1985). The variability seen is probably due at least in part to
proteolytic modification but the data indicate a subunit molecular weight
of around 110,000-120,000.
Estimated molecular weights of the holoenzyme vary from ca. 200,000
for spinach (Notton and Hewitt, 1979), tobacco (Mendel and Miiller, 1980),
barley cv. Golden Promise (Small and Wray, 1980) and wheat (Jones and
Ni Mhuimhneachain, 1985) through 220,000 or 230,000 for barley cv.
Steptoe (Kuo et aI., 1980) or squash (Redinbaugh and Campbell, 1985) to
270,000 for spinach (Nakagawa et al., 1985). This suggests that the native
enzyme is a dimer.
Although the higher plant enzyme is usually assumed to be a homo-
dimer there is no definitive biochemical evidence for this. Pan and Nason
(1978) reported that the Neurospora nitrate reductase had subunits of
unequal size (115,000 and 130,000 molecular weight), similar to the results
from spinach. However, both protein species had the same N-terminal
amino acid (glutamate) and similar peptide maps suggesting that they were
produced by proteolytic interconversion of one to the other and that the
enzyme was a homodimer. When attempts were made by Horner (1983) to
reduce proteolysis by shortening the purification procedure in the presence
of proteinase inhibitors, the subunit molecular weight of the N. crassa
enzyme was 145,000. Thus the 115,000 and 130,000 molecular weight
species seen by Pan and Nason (1978) are probably proteolytic derivatives
of a larger subunit. Genetic evidence from Neurospora crassa (Tomsett and
Garrett, 1980) and from Aspergillus nidulans (Cove, 1979) is consistent with
the enzyme being a homodimer.
Direct and indirect evidence from a few NRs suggests that the flavin is
flavin adenine dinucleotide (FAD) (Hewitt and Notton, 1980; Redinbaugh
and Campbell, 1985). The activity of several NRs is stimulated by
exogenous FAD (Maretski et al., 1967; Schrader et al., 1968). suggesting
that in some cases it is readily dissociable and, in these instances at least,
not covalently bound. Purified NR from spinach (Notton et al., 1977),
tobacco (Mendel and Miiller, 1980), barley (Somers et al., 1982) and squash
(Redinbaugh and Campbell, 1985) have spectra indicative of the presence
of a b-type cytochrome (cytochrome b m ). The first definitive evidence for
the presence of molybdenum was obtained with the spinach enzyme
(Notton and Hewitt, 1971 a) although it had previously been shown that in
barley a non-functional form of the enzyme was synthesised in the
presence of the molybdenum analogue, tungsten (Wray and Pilner, 1970).
Prosthetic group stoichiometry has been determined so far for only one
higher plant NR (from squash) and suggests the presence of one FAD, one
haem and one Mo per 115,000 molecular weight subWlit (Redinbaugh and
Campbell, 1985). It is most likely that other higher plant NRs will fit this
NAD!PlH ~ FAD ----. cyt b ~ Mo-co ----7 NO-

NAD!plH --> FAD ----> cyt bss,---+ Mo-co 3
1
cytc
Fig. 1. Schematic representation of nitrate reductase showing probable sites of
interaction of substrates and electron donors. FAD = Flavin adenine dinucleotide;
FMNH2 = Reduced flavin mononucleotide; MV = methyl viologen dye; Mo-co =
molybdenum cofactor
pattern. Electron flow from NADH is generally accepted to be via flavin

and haem to molybdenum, which acts as the terminal electron donor to
nitrate (reviewed in Hewitt and Notton, 1980) (Fig. 1).
The redox centres of other electron transport proteins, such as flavo-
cytochrome b 2 of bakers yeast (Naslin et al., 1973; Jacq and Lederer, 1974)
and the haemomolybdoprotein sulphite oxidase of rat liver (Johnson and
Rajagopalan, 1977; Southerland and Rajagopalan, 1978) are organised in
the proteins in domains (that is in independently folded, functionally intact
regions of the polypeptide chain). The domains are considered to be held
together by a flexible and loosely structured exposed region (hinge region)
which is very sensitive to proteinase attack. We have suggested that the
FAD and haem components of higher plant NR are contained in separate
functional domains of the ca. 115,000 molecular weight subunit and that an
additional domain may be involved in binding of molybdenum (Brown et
al., 1981). More recently L6 and Lederer (1983) have shown that the haem
of the N. crassa NR is contained in a domain of 10,000-12,500 molecular
weight which can be released from the native enzyme by chymotrypsin
digestion.
ii) The Molybdenum Cofactor (Mo-co)

The molybdenum of probably all molybdoenzymes, except nitrogenase
(Shah and Brill, 1977; Pienkos et al., 1977), is carried on a dissociable, dia-
lyzable and oxygen labile structure (Lee et aI., 1974; Amy and Rajago-
palan, 1979), the molybdenum-cofactor (Johnson, 1980; Johnson et al.,
1980; 1984). This conclusion derives from the earlier observation that acid
treatment of a variety of molybdoenzymes from diverse phylogenetic
sources released a component which was able to reconstitute NADPH-NR
activity in vitro from extracts of the nitrate reductase-minus mutant of Neu-
rospora crassa nit-1 (Ketchum et al., 1970; Nason et al., 1971). The Mo-co
has now been identified, on the basis of the characteristic fluorescence of
its oxidation products, as a component of the NR of Chlorella (Solo-
106 J. L. Wray
Fig. 2. Proposed structure of the molybdenum-cofactor (Johnson and Rajagopalan,

1982)
monson et al., 1984) and squash (Redinbaugh and Campbell, 1985) as well
as of Neurospora crassa (Kramer et al., 1984).
a) Structure
Structural analysis of the functional Mo-co has been impossible due to its
extreme lability when released from the protein environment. However,
chemical, mass spectral and NMR studies of its human metabolic degra-
dation product, urothione (Johnson and Rajagopalan, 1982), and of two
stable fluorescent oxidation products derived from the molybdenum
cofactor of chicken sulphite oxidase (Johnson et al., 1980, 1984) have pro-
vided information on its probable structure (Fig. 2).
The Mo-co is considered to be a complex between molybdenum and a
novel phosphorylated pterin, molybdopterin. The side chain contains at
least four carbon atoms and most probably two sulphur atoms. The pterin
presumably acts as a chelator of the metal, interfacing it to the protein and
conferring on it biological activity (Wahl et al., 1984). The nature of
molybdenum ligation is unclear but the two side-chain sulphur atoms, and
probably an oxygen atom of the phosphate group, perhaps supply at least
part of the thiol and oxo ligands of molybdenum detected in several molyb-
do enzymes by EPR (Meriwether et al., 1966; Bray, 1980) and EXAFS
studies (Cramer et al., 1981).
b) Role of Molybdenum Cofactor in Assembly of Nitrate Reductase

In addition to its role in catalysis, the Mo-co is responsible for dimerisation
of the flavohaemoprotein subunits of NR. Reconstitution of NADPH-NR
from N. crassa nit-l extract by molybdoenzyme-derived component is in
fact due to dimerisation of the monomer flavohaemoprotein subunits
present in mutant extracts (Ketchum et al., 1970). This reconstitution
process has recently been examined in detail by Kramer et al. (1984) who
showed that the molybdopterin moeity of the nit-l Mo-co is defective.
Other evidence that the molybdopterin moeity of the Mo-co is responsible
for dimerisation ability of the Mo-co comes from the observations that
demolybdo-molybdenum cofactor, produced either by molybdenum star-
vation (Hewitt et al., 1977; Gewitz et al., 1981) or mutation (Mendel et aI.,
1981) is able to effect dimerisation, as is the tungsten analogue of the
Mo-co (Wray and Filner, 1970). Dimerisation appears to involve inter-
action of the Mo-co with a protein thiol on the flavohaemoprotein subunit

(which may also function as a molybdenum ligand) since nit-l flavohaemo-
protein subunits treated with the thiol-blocker, N-ethylmaleimide, are not
dimerised by exogenous Mo-co (Wahl et al., 1984).
Recent refinements have allowed this in vitro reconstitution phenom-
enon to function as a quantitative assay for higher plant Mo-co (Mendel,
1983; Mendel et al., 1985).
iii) The Catalytic Activities Associated with the Nitrate Reductase Holo-
enzyme
In addition to the overall physiological reaction (NADH dependent nitrate
reduction) the higher plant enzymes carry a dehydrogenase (diaphorase)
function (which allows the in vitro transfer of electrons from NADH to a
variety of electron acceptors such as nitroblue tetrazolium, dichlorophenol-
indophenol and cytochrome c) as well as reduced flavin mononucleotide
(FMNH) and reduced viologen dye nitrate reductase activity. These so-
called partial activities are considered to be catalysed by specific regions of
the NR molecule (Fig. 1).
The dehydrogenase function, usually measured as cytochrome c
reductase (CR) activity, is FAD-dependent in N. crassa (Garrett and
Nason, 1969). It is not clear whether haem is obligatorily involved (Fido et
al., 1979). Removal of molybdenum (Gewitz et al., 1981) or substitution by
tungsten (Wray and Pilner, 1970; Notton and Hewitt, 1971 b) does not
affect the CR activity of NR. These results suggest that the dehydrogenase
function is catalyzed by the proximal part of the electron transport chain
and that it is an activity of the flavohaemoprotein subunit, independent of
the Mo-co. We (Brown et al., 1981; Wray and Kirk, 1981) and others
(Hamano et al., 1984a, b) have described a ca. 40,000 molecular weight CR
species which is released from barley NR after proteolytic attack by an
endogenous leupeptin-inhibited cysteine endoproteinase (Miller and
Huffaker, 1981). We have suggested that this species probably carries at
least the FAD domain together with the NADH binding site (dinucleotide
fold) (Thompson et al., 1975) of the native NR flavohaemoprotein subunit
(Brown et al., 1981). Since this species lacks haem (Campbell et al., 1984) it
may be released by proteolysis of a hinge region located between the flavin
and haem domains of the subunit.
Heat treatment at 45 for 10 min inactivates both NADH-NR and CR
activities of the barley enzyme but has little effect on FMNH-NR activity
(Wray and Pilner, 1970). Both NADH-NR and CR activity of the spinach
enzyme are protected against heat inactivation by FAD (Relimpio et al.,
1971). These results suggest that the FAD-protected, heat labile part of the
subunit (the flavin domain) is not required for the expression of
FMNH-NR activity and that electrons are donated from FMNH (and
probably from reduced viologen dye) to a later, but still undefined, site in
the electron transport chain. Since NR partial activities are molybdenum
dependent (Wray and Pilner, 1970), FMNH-NR and reduced viologen dye
NR activity are functions of the distal part of the electron transport chain.
108 1. L. Wray
Nothing is known about how the product of the NR apoprotein gene is

processed to generate the flavohaemoprotein subunit. Insertion of FAD and
haem occurs in some as yet undefined way, either before or after interaction
with the Mo-co causes dimerisation and leads to the formation of the holo-
enzyme. The coding sequence of the NR apoprotein gene is at least 3 kb in
length and represents a large target for mutagenic attack. Base substitution
or deletion may be expected to affect the expression of the activities of the
NR molecule by abolishing or altering domain function or by interfering
with domain-domain interactions. In some of the studies discussed below
attempts have been made to locate the approximate site of mutations within
the apoprotein gene by measuring the partial activities of NR.
C. Nitrite Reductase
The assimilatory nitrite reductases (ferredoxin nitrite oxidoreductase, EC
1. 7.7.1) of higher plants catalyse the six electron reduction of nitrite to
ammonium within the chloroplast (Guerrero et ai., 1981). The enzyme has
been purified from squash (Hucklesby et ai., 1967), spinach (Vega and
Kamin, 1977; Ida, 1977), barley (Serra et ai., 1982) and wheat (Small and
Gray, 1984) and is a monomer with a molecular weight of 60,000-63,000.
The spinach enzyme is best characterised and contains one tetranuclear
iron-sulphur cluster (Fe4S4) and one sirohaem per 60,000 molecular weight
peptide chain (Vega and Kamin, 1977; Lancaster et aI., 1979). Sirohaem,
dimethylurotetrahydroporphyrin, is an iron tetrahydroporphyrin of the
isobacteriochlorin type. Mossbauer spectroscopy of the oxidised NiR
reveals the presence of exchange interactions between the ferrihaem and
the S = OFe4S4 cluster indicating that the two centres are chemically linked
(Wilkerson et ai., 1983). This suggests that the nitrate reducing centre of the
spinach NiR is the coupled Fe4S4/sirohaem pair.
Nitrite reductase of oat has been shown to be nuclear-encoded (Heath-
Pagliuso et ai., 1984). The 60,500 molecular weight NiR of wheat is synthe-
sised on cytoplasmic ribosomes as a 64,000 molecular weight precursor
(Small and Gray, 1984). Several other chloroplast-located prote"ins have
been shown to be synthesised as higher molecular weight forms in the cyto-
plasm before transfer into the chloroplast where they are processed to the
mature form (Highfield and Ellis, 1978; Grossman et ai., 1982).
III. Genetics of Nitrate Assimilation
A. Introduction
Mutants altered in the nitrate assimilation pathway have been isolated at
the level of both the cell/protoplast and the intact plant. The latter
procedure directly gives whole mutant plants which are usually fertile, and
overcomes the problems which sometimes exist in the regeneration of
plants from mutant cell lines.
Two approaches have been used to identify NR-defective individuals

within populations of mutagenised cells/protoplasts or whole plants. The
first involves a non-selective total isolation procedure and has been used to
isolate NR-defective mutants at the protoplast level in Hyoscyamus muticus
(Strauss et ai., 1981) and at the whole plant level in barley (Warner et ai.,
1977). The other approach uses chlorate as a selective agent, based on the
suggestion of Aberg (1947) that chlorate toxicity in wheat is due to the
NR-catalysed reduction of chlorate to toxic chlorite; individuals which, for
whatever reason, lack a catalytically functional NR would be expected to
be chlorate resistant and could be selected as such.
There is abundant evidence to support Aberg's suggestion that
chlorate toxicity is mediated through the catalytic action of NR. Thus,
urea grown wheat plants (expected to have little or no NR activity) are
less sensitive to chlorate than are nitrate-grown plants. Light, which
increases NR activity, also increases chlorate sensitivity (Liljestrom and
Aberg, 1966). In soybean, chlorate sensitivity of leaf parts correlates
with their NR activity (Weaver, 1942; Harper, 1981 a). More direct evi-
dence comes from the demonstration that Chiorella (Solomons on and
Vennesland, 1972; Vega, 1972), tomato (Hofstra, )977) and tobacco
(Zabala and Filner, 1980) NRs can use chlorate as a substr:ate. However
Aberg's suggestion is not consistent with the wide-spread occurrence of
chlorate-resistant, higher plant cell-lines which still possess NR activity
(see for example Muller and Grafe, 1978; Murphy and Imbrie, 1981;
Buchanan and Wray, 1982).
Further evidence that the basis of chlorate resistance may not be as
straightforward as originally thought comes from Aspergillus where
chlorate-sensitive mutants lacking NR activity have been reported (Cove,
1976 a). In this organism both NR and the product of the regulatory gene,
nirA, are involved in the catabolism of some nitrogen compounds and also
in the mediation of chlorate toxicity. Cove has suggested that chlorate is
toxic because it mimics nitrate in mediating, via NR and the nirA gene
product, a shut-down in nitrogen catabolism. Since chlorate cannot act as a
nitrogen source, nitrogen starvation ensues. Thus in this organism at least
chlorate toxicity may not be mediated by the catalytic function of NR.
Whatever the mechanism of chlorate resistance may be, chlorate has
proved useful as a selective agent in the isolation of NR-defective mutants.
Its first reported use in higher plants was in the early 1970s (Oostindier-
Braaksma, 1970; Laan et ai., 1971; Oostindier-Braaksma and Feenstra,
1972).
Chlorate also acts as a nitrate analogue at the level of uptake.
Nitrate competes with chlorate for uptake into Arabidopsis thaliana
(Doddema and Telkamp, 1979). In barley, depletion of chlorate or
nitrate from uptake media over 2-6 h by seedlings was found to be
dependent on combined nitrate plus chlorate concentration, and total
anion uptake was equivalent at different nitrate to chlorate ratios. Line-
weaver-Burk plots of the interaction between nitrate and chlorate were
characteristic of competitive inhibition (Deane-Drummond and Glass,
110 J. L. Wray
1983). One might therefore expect that some mutations within nitrate
uptake could lead to chlorate resistance.
Higher plant nitrate assimilation mutants have been analysed using
both genetic and biochemical approaches. Conventional genetic analysis
by crossing fertile mutant plants has allowed the testing of allelism, as well
as determination of the recessive or dominant character of the mutation
and the type of inheritance. In those cases where only mutant cell lines are
available more limited genetic analysis has been carried out by comple-
mentation in somatic hybrids. It should be emphasized however that defin-
itive allocation of a function to a locus requires analysis of more than one
or a few mutant alleles at that locus.
B. Nitrate Uptake Mutations

In Aspergillus, mutations at the crnA locus lead to reduced net uptake of
nitrate into conidiophores and young mycelia. Loss of function confers
resistance to chlorate and bromate without any obvious nutritional
impairment (Tomsett and Cove, 1979; Brownlee and Arst, 1983). Only two
putative nitrate uptake mutants, isolated on the basis of chlorate resistance,
have been described from higher plants. These are the whole plant allelic
mutants Bland B 3 of Arabidopsis thaliana (Oostindier-Braaksma and
Feenstra, 1973). The mutation is monogenic and recessive and the gene
locus has been designated chll.
Net uptake of nitrate in wild-type Arabidopsis is biphasic, each phase
showing Michaelis-Menten kinetics. Phase I has a Km of 0.04 mM and
phase II a Km of 25 mM for nitrate. In the B 1 mutant, phase II does not
follow Michaelis-Menten kinetics and the net uptake of nitrate in the phase
II concentration range is considerably lower than the wild type. It is inter-
esting that this defect in phase II confers sufficient resistance to 13 mM
chlorate to allow B 1 to be distinguished from wild-type plants (Doddema
and Telkamp, 1979). B 1 is still able to take up chlorate (Oostindier-Bra-
aksma and Feenstra, 1973) as well as nitrate. It would be useful to know
whether the ability to select nitrate uptake mutants is dependent on the
concentration of chlorate used in the screen. It might be that at higher
chlorate levels mutation in one of the uptake systems would be insufficient
to allow selection at the whole plant level.
Since the phase I uptake mechanism is unaffected in B 1, Doddema and
Telkamp (1979) have concluded that uptake of nitrate in Arabidopsis is
mediated by (at least) two independent uptake mechanisms. If this biphasic
uptake mechanism is common in higher plants then it is unlikely that
mutants completely defective in nitrate uptake could be selected at the
whole plant level in a one-step chlorate screen. Individuals within M2 pop-
ulations simultaneously defective in both (or more than two) uptake genes
are likely to be extremely rare.
C. Nitrate Reductase Mutations
i) Apoprotein Gene Mutants

Extensive genetic analysis in lower eukaryotes, particularly in Aspergillus
nidulans (reviewed by Cove, 1979) but also in Neurospora crassa (Sorger
and Giles, 1965; Tomsett and Garrett, 1980) and Penicillium chrysogenum
(Birkett and Rowlands, 1981), indicates the presence of a single gene locus
coding for the apoprotein of NADPH-NR (EC 1.6.6.3) and suggests the
enzyme is a homodimer. This latter conclusion is supported by studies on
the enzyme purified from Neurospora (Pan and Nason, 1978). The situation
is rather more complicated in higher plants due to the presence of both
constitutive and nitrate-inducible forms of NR and of species which are
either NADH-specific (EC 1.6.6.1) or bispecific for NADH and NADPH
(EC 1.6.6.2) as electron donor. The isolation and characterisation of apo-
protein gene mutants in higher plants has contributed greatly to our under-
standing of the relationships between these different types and has also
stimulated studies into the characterisation of the forms of NR present in
wild-type plants.
A wide range of NR mutants having properties consistent with defects
in an apoprotein gene locus have been isolated in higher plants (Table 1).
The evidence that these are indeed apoprotein gene mutants is stronger in
some cases, for example in Nicotiana species and Hordeum vulgare, than in
others where interpretation rests solely on the demonstration of the re-
tention of xanthine dehydrogenase activity. In N. tabacum and Hordeum
vulgare the evidence is particularly strong that the apoprotein of
NADH-NR is coded for by a single gene, indicating that the enzyme is a
homodimer. Similar conclusions for other higher plant NRs are weakened
by the relatively small number of mutants which have been isolated and
characterised.
a) Nicotiana tabacum
Detailed biochemical and genetic analysis has been carried out in Nico-
tiana spp., particularly with the allotetraploid N. tabacum L. var Gaters-
leben (Muller and Grafe, 1978). NR-minus mutants were isolated by
screening mutagenised amphihaploid (n = 24) cell suspensions for
resistance to 20 mM potassium chlorate. Thirty six lines retaining xanthine
dehydrogenase activity (Mendel and Muller, 1976; Muller and Grafe,
1978; Mendel and Muller, 1979; Muller, 1983; pers. comm.) were desig-
nated nia in accordance with the gene symbol used in Aspergillus for those
mutants which were NR-minus but had xanthine dehydrogenase activity
(Pateman et al., 1964; Cove and Pateman, 1969). All these mutants had lost
the CR activity catalysed by the flavohaemoprotein subunit of NR and all
except three had lost all three (NADH-, FMNH-, BVH-) NR activities. The
exceptions are mutants Nia95, Nia26 and Nia28 which, although lacking
CR activity, still show inducible FMNH and BVH-NR activities. These
biochemical data have been interpreted to show that the nia mutants are
altered in the gene coding for the apoprotein of NR. The differing enzy-
tv
-
Table I. Higher plant nitrate assimilation mutants
Sexual Trans- Isolation

Mutant
Species Gene Function Phenotype mission to Seed Pro- Reference
Number
Progeny cedure'
Arabidopsis ch12 B2-1 ? Deficient in NR and XDH; poor yes Braaksma and
thaliana (chromo- growth on nitrate; chlorate Feenstra,1982a
some 2) resistant
chl3 B29 ? Deficient in NR and XDH; yes as above
(chromo- B31-2, B33, strong growth on nitrate;
some I) B35 chlorate resistant
~
(four
alleles) r
rgn B25 Molybdenum Deficient in NR and XDH; poor yes as above
(chromo- cofactor growth on nitrate; chlorate ~
~
some I) resistant '<

cnx B73 Molybdenum Deficient in NR and XD H; poor yes as above
(chromo- cofactor growth on nitrate; chlorate
some 5) resistant
nd B31-1 ? Deficient in NR; strong growth yes as above
on nitrate; chlorate resistant
chll BI, B3 Nitrate uptake Decreased uptake of nitrate; yes as above
(two alleles) gene retains NR and XDH; chlorate
resistant
nd B36 ? Deficient in NR; strong growth yes as above
nd B40 ? Deficient in NR; strong growth yes as above
Species Gene Mutant Function Phenotype mission to Seed Pro- Reference
Number Progeny cedure'
Glycine max nrj LNR-2 Possible apo- Constitutive-NR minus: (ace- yes 2 Nelson et aI.,
L. Merr LNR-3 and protein or regu- taldehyde oxime-evolution 1983, 1984, and
cv. Williams LNR-4 latory gene minus) retains XDH and indu- Ryan et aI., 1983
(three cible NR; growth on nitrate;
alleles) chlorate resistant
Hordeum vulgare narl 9 Alleles Apoprotein gene Deficient in NADH-NR; retains yes 3 Warner et al., a
(1)
cv. Steptoe initially XDH;growth 1977; Kleinhofs :l
(1)
coded on nitrate et al., 1980 ....
Az 1,2 etc.
o
V>
0-,
nar2 Az34 Molybdenum Deficient in NR and XDH; yes 3 Kleinhofs et al.,
cofactor reduced growth 1980
~
....
on nitrate ..,
I\)
nar4 Az72 Molybdenum Deficient in NR and XDH; yes 3 Kleinhofs et al., ....
(1)
cofactor reduced growth 1985
~
on nitrate V>
V>
cv. Winer narl Xn024 Apoprotein gene Deficient in NR; chlorate yes 4 Tokarevand
resistant, retains XD H; growth Shumny,1977; :
~
....
on nitrate Shumnyand
Tokarev, 1982; o
:l
Kleinhofs et al.,
1983; Somers et
al., 1983
nar3 Xno 18, Molybdenum Deficient in NR and XDH; yes 4 as above
Xno 19 cofactor chlorate resistant, growth on
(two alleles) nitrate
cv. Maris nar2 R9201, Molybdenum NR and XDH minus; chlorate yes 5 Bright et al.,
Mink R9401 cofactor resistant; 1983; Steven,
(two alleles) no growth on nitrate; unpublished;
conditional lethal Wray et aI., 1985
....
w
-"'"
Species Gene Mutant Function Phenotype mission to Seed Pro-
Number Reference
Progeny cedurea
nd 11301 Molybdenum as above yes 5 as above

cofactor
12202 Molybdenum as above yes 5 as above
cofactor
Hyoscyamus cnxA MA-2 Molybdenum NR and XDH minus; no 6 Strauss et at.,
muticus cofactor no growth on nitrate, chlorate 1981; Lazar et
resistant; at., 1983
Mo repairable
O,Q Molybdenum as above no 7 H. Fankhauser,
cofactor pers. comm.
nd Molybdenum as above no 7 as above :-<
(comple- cofactor r
mentation
group B) ~
III
nd 12D 12 Molybdenum NR and XDH minus; no 6 Strauss et at., '<
(comple- cofactor no growth on nitrate; chlorate 1981; Fank-
mentation resistant; hauser et at.,
group C) Mo repairable 1984
nd XIVE9 Molybdenum NR and XDH minus; no 6 Fankhauser et
cofactor no growth on nitrate; chlorate at., 1984; Strauss
resistant et at., 1981
T,C as above 7 H. Fankhauser,
(comple- pers. comm.
mentation
group D)
nd VIC2 Molybdenum as above no 6 Strauss et ai.,
(comple- cofactor 1981; Fank-
mentation hauser et at.,
group E) 1984
Species Gene Mutant mission to Seed Pro- Reference
Function Phenotype
Number Progeny cedurea
Nicotiana ptumba- nia NA (7 lines) Apoprotein gene NR minus; retains XDH; no no 8 Marton et at.,
ginifolia growth on nitrate; chlorate 1982a, b
resistant
26 lines Apoprotein gene NR minus; retains XD H; no yes 9 Negrutiu et ai.,
growth on nitrate; chlorate 1983
a
~
resistant ~
=
....
....
cnxA NX1, NX9 Molybdenum NR and XDH minus; no 8 Marton et ai., 0
en
(two alleles) cofactor no growth on nitrate; chlorate 1982 a, b; Xuan 0
resistant; et ai., 1983 ...,
Mo repairable ~
CNX20 Molybdenum NRand XDH minus; yes 9 Negrutiu et ai.,
....
'"I
CNX82 cofactor no growth on nitrate; chlorate 1983; Dirks ~

~
(two alleles) resistant; et ai., 1985 -

Mo repairable '"en
cnxB NX24 Molybdenum NR and XDH minus; no 8 Marton et ai., ~:
cofactor no growth on nitrate; chlorate 1982 a, b
resistant
PI....
0
CNX27 Molybdenum NR and XDH minus; no 9 Negrutiu et aI.,
cofactor no growth on nitrate; chlorate 1983; Dirks
=
resistant et ai., 1985
cnxe NX21 Molybdenum NR and XDH minus; no 8 Marton et ai.,
cofactor no growth on nitrate; chlorate 1982 a, b
resistant
cnxD CNXI03 Molybdenum NRand XDH minus; yes 9 Negrutiu et ai.,
cofactor no growth on nitrate; chlorate 1983; Dirks
resistant et at., 1985
VI
-
Sexual Trans- Isolation 0'1
Species Mutant Function Phenotype
--
Gene mission to Seed Pro- Reference
Number Progeny cedurea
Nicotiana tabacum nia I nia2 36 alleles Duplicate apo- NR-minus; no growth on yes-fertile 10 Muller and
cv. Gatersleben protein genes nitrate; chlorate resistant, retains plants regen- Grafe, 1978;
XDH erated from 15 Muller, 1983
cell lines
cnxA I cnxA 2 Cnx 68 Duplicate NR and XDH minus; yes-fertile 10 Muller and
Cnx 101 molybdenum no growth on nitrate; chlorate plants regen- Grafe, 1978;
Cnx109 cofactor genes resistant; erated from Grafe and
Cnx 135 Mo repairable Cnx135 Muller, 1983;
(four A. Muller, pers.
alleles) comm.
cv. Xanthi cnxB 042, PI2 Molybdenum NR and XDH minus, no II Buchanan and
P31, P47 cofactor no growth on nitrate; chlorate Wray, 1982; :-0
(four resistant Xuan etal., r-'
alleles) 1983; Mendel et
aI., 1984 ~
III
cv. Xanthi nd 'clr 19 Apoprotein gene? NR minus, retains XDH; no no 12 Evola, 1983 a, b '<
growth on nitrate; chlorate
resistant
clrO Molybdenum NR and XDH minus; no 13 as above
cofactor no growth on nitrate; chlorate
resistant
Petunia hybrida nd line I Possible apo- NR minus; no growth on nitrate; no 14 Steffen and
var. Mitchell protein gene chlorate resistant; retains XDH Schieder, 1984
nd line 2 Molybdenum NR and XDH minus; no 14 as above
cofactor no growth on nitrate; chlorate
resistant
nd lines 3 and Molybdenum NR and XDH minus; no 14 as above
4 cofactor no growth on nitrate; chlorate
resistant;
Mo repairable
Mutant Sexual Trans- Isolation
Species Gene Function Phenotype mission to Seed Pro- Reference
Number
Progeny cedure a
Pisum sativum cv. narl A317, A334 Possible apo- Deficient in NR yes 15 Kleinhofs et al.,
Juneau (two alleles) protein gene 1978; Warner et
aI., 1982
nar2 A300 Molybdenum Deficient in NR and XDH yes 15 as above
cofactor
cv. Rondo nd El Molybdenum Deficient in NR and XDH; yes 16 Feenstra and Cl
t1>
cofactor chlorate resistant Jacpbsen,1980; ::st1>
Jacobsen et al., ......
(S.
1984 rI>
o-,
NR-deficient mutants have also been reported in Datura innoxia (King and Khanna, 1980) and Rosa damascena (Murphy and Imbrie, 1981) cell cul-
tures. ENU, N-ethyl-N-nitrosourea; EMS, ethyl methane sulphonate; MNNG, N-methyl-N'-nitro-N-nitrosoguanidine, nd - not designated. ......
.,~
aIsolation Procedures ~
t1>
1 6 day old M 2plants (40 mM EMS in M I) resistant to 13 mM N aCI0 3 ;l>
2. 10-12 day old plants derived from seed mutagenised for four successive generations with a variety of chemical and physical mutagens and rI>
rI>
screened for decreased visual damage in the presence of 0.1 mM KCIO l . Rescreened for lowered in vivo NR.
3 7 day old M2 plants (I mM azide pH3 for 2h in M I) screened for lowered in vivo NR activity.
:
4 7 day old M 2plants (0.25 % EMS for 10-12 h in M I) screened for resistance to 8 mM KCIO l for 2 days. ~
5 6 day old M 2plants (I mM azide pH 3 for 2 h in M I) screened for resistance to 10 mM KCIO l for 9 days.
o::s
6 Haploid mesophyll protoplasts mutagenised with 20 mg. I-I MNNG and screened for amino acid auxotrophy followed by nitrate non-utilization.
7 Haploid mesophyll protoplasts mutagenised and screened for resistance to chlorate.
8 Haploid protoplasts mutagenised with either a 60CO source (0.04 Gy. sec-I) or ENU and screened for resistance to 40mM KCIO l .
9 Haploid protoplasts screened for spontaneous resistance to 125 mM KCIO l .
10 Amphihaploid cell suspension mutagenised with 0.25 mM ENU and screened for resistance to 20 mM KCIO l .
II Amphihaploid cell suspension mutagenised with 0.4 % EMS for I h and screened for resistance to 20mM KCIO l .
12 Amphihaploid cell suspension mutagenised with 0.5mM ENU and screened for resistance to 20mM KCIO l .
13 Amphihaploid cell suspension screened for spontaneous resistance to 20 mM KCIO l .
14 Cell colonies derived from haploid mesophyll protoplasts mutagenised with X-rays (1000 R) screened for resistance to 100 mM KCIO l .
15 M2 plants (I mM azide pH3 for 2h in M I) screened for lowered in vivo NR activity. ......
......
16 12 day old M2 plants (0.3 % EMS for 4h in M I) screened for resistance to 20mM KCIO l . -...I
118 J. L. Wray
matic activities possessed by mutants may be due to point mutations at dif-

ferent places within the nia gene locus.
Fifteen of the thirtysix nia mutants have been regenerated to fertile
amphidiploid plants and have been analysed genetically through sexual
crosses. As anticipated from the biochemical evidence, all fifteen mutants
proved allelic (Muller, 1983). Segregation among F2 progeny from crosses
between mutant and wild-type plants and among test-cross progeny
showed the regenerated plants to be homozygous, double mutants. The
NR-minus phenotype is conferred by two unlinked recessive nuclear muta-
tions which define a pair of duplicate loci (nia1 nia2).
This interesting finding may be explained by the fact that N. tabacum
(2 n = 48) is believed to have arisen by hybridisation between two diploid
progenitor species N. sylvestris (2 n = 24) and N. tomentosiformis (2 n = 24)
(Gray et al., 1974). One of the NR apoprotein gene loci is presumably
derived from each parent. Genetic analysis showed that both loci (nial and
nia2) mutated after initiation of the parental cell culture, most probably
after treatment with mutagen, since mutants were recovered only after
mutagen treatment. The low frequency of recovered NR-minus mutants
(about 10- 7) is consistent with simultaneous induction of two independent
mutations and is much higher (10- 3 to 10- 4) in similar work on monoploid
cells of the true haploid Nicotiana species, N. plumbaginifolia (Marton et
al., 1982 a) discussed below. This shows that both homeologous gene loci
are functional within N. tabacum and raises the possibility that two types of
apoprotein are synthesised. Random assembly of flavohaemoprotein sub-
units might generate three types of NADH-NR. Since constitutive NR
activity as well as nitrate inducible NR activity is missing from nia mutants
(Muller and Grafe, 1978; Muller, 1983) it is likely that their apoproteins are
the products of the same gene.
Construction of mutant plants carrying different numbers (up to four)
of nia+ genes, single mutants and even plants possessing only one
wild-type allele at either of the nia loci were all normal with respect to
growth on nitrate, growth on soil and NR level at most developmental
stages. Only at the early seedling stage did the NR activity respond to the
number of nia genes, as shown by the in vivo NR activity and chlorate sen-
sitivity. Thus the initially constitutive expression of the nia+ genes changes
during seedling development to a strictly regulatory type of gene
expression (Muller, 1983).
A further NR-minus chlorate resistant mutant has been isolated from
mutagenised protoplasts of a dihaploid plant of N. tabacum L. var Xanthi
(Evola, 1983 a). This mutant, clr19, which retains xanthine dehydrogenase
activity, lacks nitrate-inducible CR activity but possesses a very low level
of nitrate inducible BVH-NR activity, suggesting that it might be an apo-
protein gene mutant altered in the proximal part of the electron transport
chain. There is no further biochemical information on this mutant and,
although clr19 scions flowered when grafted on wild-type stocks it is not
known whether they are allelic to Gatersleben mutants.
b) Nicotiana plumbaginifolia
Putative NR-minus apoprotein gene mutants have been isolated from the
true haploid Nicotiana species N. plumbaginifolia (n = 10). Marton et al.
(1982a) isolated 36 chlorate-resistant clones of which 29 were fully defi-
cient in NR. Of nine clones examined, five (designated NA) possessed xan-
thine dehydrogenase activity and were allelic as shown by non-complemen-
tation in somatic hybrids (Marton et al., 1982b). Attempts to reconstitute
NR activity by cohomogenisation of cells of each of the five lines with the
apoprotein gene mutant, Nia 63, of N. tabacum var. Gatersleben were
unsuccessful, suggesting that they may also be apoprotein gene mutants
(Marton et al., 1982 a). In N. plumbaginifolia selection for resistance of
haploid protoplasts to potassium chlorate (40 mM) produced spontaneous
mutants at a frequency of 10- 5 to 10- 6 (Negrutiu et al., 1983). In Aspergillus,
nia mutants selected via chlorate resistance without prior mutagenesis are
more likely to be deletion mutants (Cove, 1976 b). Twenty six lines which
were NR-minus but retained xanthine dehydrogenase activity were clas-
sified as apoprotein gene mutants (Negrutiu et al., 1983). Plants regen-
erated from twelve of these lines were shown to express and transmit the
NR-minus phenotype to the progeny in a Mendelian fashion. The plants
were allelic and diploid and the mutation was inherited as a single
recessive nuclear gene (Dirks et al., 1985).
c) Petunia hybrida
Steffen and Schieder (1984) have recently reported the isolation of an NR-
minus chlorate-resistant (lOOmM) line of Petunia hybrida var. Mitchell.
The line retained xanthine dehydrogenase activity and was tentatively
classed as an apoprotein gene mutant. Plants could not be regenerated.
d) Hordeum vulgare
Whole plant mutants have been directly isolated in barley, Hordeum
vulgare L. cv. Steptoe (Kleinhofs et al., 1978). Mutagenesis in the M j was
carried out with sodium azide at low pH. Kleinhofs and coworkers initially
attempted to screen for loss of NR activity in M2 seedlings through the use
of chlorate. Although chlorate resistant seedlings were identified at a fre-
quency of 6 per 10,000, all appeared lethal and attempts to transplant and
propagate these seedlings failed.
Subsequently NR mutants were identified by a rapid semiquantitative
in vivo NR assay of individual seedling leaves. Seedlings showing 10 % or
less of control NR activity were classified as being NR defective (Warner et
al., 1977). Nine NR-deficient mutants were shown to be allelic when tested
by reciprocal crossing in all possible combinations and the locus repre-
sented by these mutants (Az 12, 13, 23, 28, 29, 30, 31, 32 and 33, subse-
quently renumbered a, ... i) was designated narl. The narl gene is a
nuclear gene (Kleinhofs et al., 1980). These mutants have NR activities
ranging between 2 and 7 % of the wild type when assayed in vitro with
NADH as electron donor (Somers et al., 1983).
120 J. L. Wray
Several lines of evidence together show unequivocably that the narl

locus is the apoprotein gene locus. Some nar1 alleles possess no nitrate
inducible CR activity (carried by the proximal part of the electron
transport chain), some have levels similar to wild type, whilst one (nar1d)
has an extremely high level. All mutants except one (nar1h) lack nitrate-
inducible FMNH-NR activity, catalysed by the distal end of the electron
transport chain (Kleinhofs et ai., 1980). NR CRM levels correlated with the
level of NR-associated enzyme activities (Kuo et ai., 1981; Somers et ai.,
1983). These results are consistent with the suggestion that the alleles are
due to mutations at different sites within the nar1 locus. These mutations
produce mutant flavohaemoprotein subunits which are responsible for the
different levels and types of subunit-associated enzyme activity (Kleinhofs
et ai., 1980). Cleveland mapping of the nar1d CR species shows that it con-
tains a glutamic acid substitution when compared to the wild-type NR
(Kuo et ai., 1984). Further evidence that the flavohaemoprotein subunit is
defective in nar1 alleles comes from the observation that functional Mo-co,
released from bovine xanthine oxidase by heat treatment, cannot recon-
stitute NR activity in vitro from extracts of any of the nar1 alleles (Na-
rayanan et ai., 1984). A nar1 allele Xn024, designated nar1 j (Kleinhofs et
ai., 1983), has also been isolated from the barley cultivar Winer (Shumny
and Tokarev, 1982).
When grown to maturity with nitrate as sole nitrogen source, the nar1
alleles nar1a and nar1b possessed as much dry weight and reduced
nitrogen per plant as the wild-type control. Similar results were obtained
with aseptically grown excised embryos (Warner and Kleinhofs, 1981).
When grown in the field the total vegetative dry weight, total reduced
nitrogen and percent grain protein of the mutants was the same as the wild
type, whereas grain yield was lower (Oh et ai., 1980). These results indicate
that nar1 alleles with NR levels less than 10 % of wild type can utilise
nitrate as sole nitrogen source and perform almost as well as wild-type
plants.
It might be thought that the ability of these nar1 alleles to utilise nitrate
was due to residual NADH-NR activity present as a result of leaky muta-
tions within the nar1 locus. However, very surprisingly, the activity present
in these alleles appears to be that of a bispecific NAD(P)H-linked enzyme
which cannot be detected in wild-type plants (Dailey et ai., 1982a, b). This
bispecific NAD(P)H-NR cross-reacted with antiserum raised against the
wild-type NADH-NR but with a much lower specificity, perhaps due in
part to its slightly smaller subunit molecular weight (A. Kleinhofs, pers.
comm.). Although the NAD(P)H-NR plays a significant role in nitrate
reduction in narl mutants, the mechanism which allows it to be expressed
in these mutants is not clear.
e) Pisum sativum
Two whole plant mutants have been isolated in pea, Pisum sativum, by
screening in the M2 for low in vivo NR activity (Kleinhofs et at., 1978;
Warner et al., 1982). The mutant plants (A317 and A334) had less than 6 %
of the wild-type in vitro NADH-NR and FMNH-NR activity but still
retained inducible NADH-CR activity. The mutations of A317 and A334
are monogenic, allelic, express incomplete dominance and have been
designated nar!' The mutants possess xanthine dehydrogenase activity.
The most likely interpretation of these results is that the narl locus is the
apoprotein gene locus. Mutations in both alleles appear to be associated
with the distal end of the electron transport chain since FMNH-NR
activity is impaired whilst the NADH-CR activity associated with the
proximal end of the electron transport chain is not. Nitrite reductase
activity was inducible, as in the wild type, but activities were higher,
perhaps reflecting the much higher nitrate levels which accumulated in the
mutants (Warner et al., 1982).
f) Glycine max
Mutant lines of soybean have been invaluable in understanding the rela-
tionship between the different forms of NR which exist in this species
(Evans and Nason, 1953; Jolly et al., 1976; Campbell, 1976; Lahav et al.,
1976; Kakefuda et al., 1983). Whole plant mutants were obtained by
screening M2 seed, derived from soybean (Glycine max L. Merr. cv. Wil-
liams) seed which had been treated for four successive generations with
various chemical and physical mutagens (designated the M, seed) for
resistance to 0.1 mM potassium chlorate. Chlorate toxicity was assessed by
necrosis of the cotyledon margins and/or chlorosis and necrosis of the
unifoliate leaves with subsequent stunted leaf expansion. Forty-nine po-
tential NR-deficient plants were selected from 12,000 M2 seedlings. Thirty-
eight of these produced seed. The M3 plants were subjected to a second
chlorate screen and leaf in vivo NR activity was determined in resistant
plants. Selected M3 plants were harvested and three lines with decreased
NR activity (LNR-2, LNR-3 and LNR-4) were identified in the M4 (Nelson
et al., 1983). The three lines are allelic (Ryan et al., 1983).
Wild-type soybean possesses two main types of NR, a form present in
urea-grown plants which has been called constitutive since it does not
require nitrate for expression, and a form which is nitrate inducible. In
young soybean leaves grown on nitrate the constitutive NR makes up
approximately 50 % of the total activity. The NR activities of the mutant
lines are approximately 50 % of the wild-type activity (assayed in vivo and
in vitro with NADH) when grown on nitrate as sole nitrogen source, but
NR activity was absent from leaves of the urea-grown mutant. Thus the
mutants lack the constitutive NR activity but still retain the nitrate induc-
ible NR activity (Nelson et al., 1983). The nitrate inducible NR was
purified from nitrate-grown mutant plants and shown to be a NADH-NR
(EC 1.6.6.1) with a pH optimum of 7.5 (Streit et al., 1985). This enzyme,
which is present in all other plant species examined, had not previously
been identified in soybean due to the presence of the constitutive NR. The
nitrate grown mutant is a source of pure soybean NADH-NR.
122 J. L. Wray
The urea-grown soybean mutant LNR-2 (designated nr]) had also lost
in vitro FMNH-NR, and CR activity was much reduced. Since the xanthine
dehydrogenase activity of nr] was unaffected, Nelson et aI., (1984) sug-
gested that the mutation in nr] was in the apoprotein gene of the consti-
tutive NR. Evolution of acetaldehyde oxime which occurs in wild-type
soybean plants under anaerobic conditions (Harper, 1981 b; Mulvaney and
Hageman, 1984) did not occur in the nr] mutant, indicating a role for the
constitutive NR in gas evolution. Absence of constitutive NR and evo-
lution of acetaldehyde oxime are controlled by a single recessive nuclear
gene (Ryan et al., 1983). The absence of constitutive NR activity in the
mutants did not increase nitrate levels or decrease reduced-nitrogen con-
centrations in the plants. The ability of these soybean mutants to maintain
apparently normal nitrogen metabolism, despite lowered NR activity, is
similar to the narl barley mutants (Warner and Kleinhofs, 1981).
Immunological and blue-dextran Sepharose chromatographic studies
have subsequently shown the constitutive NR to consist of two species
(Robin et al., 1985; Streit et al., 1985). One form, eluted with NADPH from
blue-Dextran Sepharose, is more active with NADPH than NADH as
electron donor, has a relatively high Km for nitrate and a pH optimum of
6.5 and is immunologically identical to the N AD (P) H -NR described by
Jolly et al. (1976) and Campbell (1976). The other form, eluted with
NADH, is more active with NADH, has a pH optimum of 6.5 and is the
NADH-NR described by Jolly et al. (1976). The nitrate-inducible NR has a
sedimentation coefficient of 7.6 like most other NADH-NRs examined.
The constitutive NRs had sedimentation coefficients of 5.6 (NADPH-
eluted) and 6.0 (NADH-eluted) and had higher mobilities on polyacryl-
amide gel than the nitrate-inducible NADH-NR.
Since loss of constitutive NR activity is due to a mutation in a single
recessive nuclear gene, the expression of both forms of constitutive NR
must be controlled by the same gene. This would imply that they share
common flavohaemoprotein subunits or that the nr] mutation is in a regu-
latory, rather than a structural, gene locus. The functions of these consti-
tutive NRs are unknown but they are active in vivo (Ryan et al., 1983).
g) Arabidopsis thaliana
Whole plant mutants have been isolated by screening M z plants for
resistance to 13 mM chlorate (Braaksma and Feenstra, 1982a). Whilst the
genetic analysis of these mutants is the most sophisticated so far carried
out in higher plants, the nature of some of the mutations is unclear.
Braaksma and Feenstra (1982a) have argued that the mutations in chl2
and chl3 probably do not affect the Mo-co and that both are apoprotein gene
mutations. Since Arabidopsis apparently possesses only NADH-NR this
would suggest that the situation in Arabidopsis is different from all other
eukaryotic NR systems where biochemical and genetic analysis indicates the
presence of only one type of apoprotein subunit in individual NR species.
The argument that ch12 is an apoprotein gene locus is based solely on
the observation that the 8 S NR/CR peak seen after sucrose gradient
analysis of ch12 extracts has a lower ratio of NR to CR activity than is seen
in the wild type, indicating perhaps some lesion in the apoprotein which
affects CR activity. However, xanthine dehydrogenase activity in chl2 is
also reduced to less than half that of the wild type, suggesting a defect in
the Mo-co.
Their argument that chl3 is also altered in an apoprotein gene is based
on the observation that there are irregularities in the aggregation of the
enzyme complex and that this might be caused by an altered flavohaemo-
protein subunit. Whilst this is a possibility, it is known that aggregation is
also a function of Mo-co structure and does not explain why xanthine
dehydrogenase levels in this mutant are also lowered compared to the wild
type. Thus it is not clear whether either ch12 or chl3 represent apoprotein
gene mutations. At least part of the problem may be attributed to the leaky
nature of the mutants. ch12 and chl3 possess approximately 10 % and 20 %
of the wild-type NR activity respectively.
ii) Molybdenum Cofactor Mutants
Mutations within Mo-co genes were first described in the lower eukaryote,
Aspergillus nidulans (reviewed in Cove, 1979). Some mutants selected for
their inabililty to utilise nitrate as a nitrogen source were also unable to
utilise hypoxanthine (Pateman et al., 1964). Both nitrate reductase and xan-
thine dehydrogenase were undetectable or present at very low levels in
these mutants and, since both were molybdoenzymes, the defect was sug-
gested to be within a molybdenum cofactor shared by these two proteins
(Cove, 1963; Pateman et al., 1964). These mutants were thus designated cnx
(cofactor for nitrate reductase and xanthine dehydrogenase).
Heterokaryon complementation tests for nitrate utilization have been
carried out for over four hundred independently-isolated cnx mutants. The
cnx mutants fall into seven complementation groups, designated A, B, C, E,
F, G and H (Cove, 1963; Rever, 1966; Hartley, 1969, 1970). cnx A, Band C
mutants show an overlapping complementation pattern and are closely
linked. cnxA and cnxC mutants may involve two distinct genes, with cnxB
mutants lacking both functions. cnxA, Band C mutations are unlinked to
cnxE, cnxF, cnxG and cnxH which are in turn unlinked to one another
(Cove, 1979). More recently mutants at a further cnx locus, cnxJ, have been
described. Arst et al., (1982) suggest that the cnxJ gene product plays a
regulatory role in Mo-co synthesis. Thus at least six and possibly seven
genes can mutate to give the cnx phenotype. Mutations within the Mo-co
have been described also in N. crassa (Sorger and Giles, 1965; Coddington,
1976; Tomsett and Garrett, 1980), Penicillium chrysogenum (Birkett and
Rowlands, 1981) and in prokaryotes, for example, Escherichia coli (Ruiz-
Herrera et al., 1969; Stewart and MacGregor, 1982).
A consideration of the structure (Fig. 2) and function of the Mo-co pro-
vides an explanation of why so many cnx loci are involved in its biosyn-
thesis and suggests that two metabolic pathways are probably required.
One pathway, leading from mainstream pterin metabolism, involves modi-
124 1. L. Wray
fication of the pterin nucleus to generate the phosphorylated pterin deriv-

ative, molybdopterin. The other pathway, some steps of which are specu-
lative, is envisaged to include molybdate uptake into the cell, intracellular
transport and storage of molybdate, generation of a form of molybdenum
which is in the correct redox and ligandable state, and finally liganding of
molybdenum to molybdopterin to generate the functional Mo-co. Whether
this last step occurs before or after interaction of molydopterin with flavo-
haemoprotein subunits is not known. Some of the cnx loci will be involved
in synthesis of the phosphorylated carbon side chain and in attachment of
the sulphur atoms of molybdopterin (it is unlikely that they will be
involved in synthesis of the pterin nucleus itself since mutations in such
loci would be expected to be lethal) and some will be involved in what may
be loosely described as molybdate metabolism/processing and in ligan ding
of molybdenum to molybdopterin.
Arst et al. (1970) have shown that high concentrations of molybdate
supplied to growing cultures of cnxE mutants partially restored NR and
xanthine dehydrogenase activity (in vivo repair) and they suggested that the
cnxE gene product might be involved in the insertion of molybdenum into
the cofactor. Since some cnxH mutations are temperature-sensitive, this
locus may specify a structural component of the Mo-co (MacDonald and
Cove, 1974) but what this might be is unclear. Evidence for a molybdate
reduction step has recently come from work with E. coli (Campbell et al.,
1985). Mutations in this step have been isolated but not yet characterised.
Mo-co mutants have been described in a wide variety of higher plants
(Table 1). Identification of mutants as cnx has depended largely on the
demonstration of the pleiotropic loss of NR and xanthine dehydrogenase
activity with the retention of a functional flavohaemoprotein subunit.
a) Nicotiana
Molybdenum cofactor mutants have been isolated in N. tabacum L. var
Gatersleben (Muller and Grafe, 1978) and var Xanthi (Buchanan and
Wray, 1982; Evola, 1983 a) and in N. plumbaginifolia (Marton etal., 1982a;
Negrutiu et al., 1983).
Four NR-minus chlorate resistant lines of N. tabacum L. var Gaters-
leben were isolated in three independent experiments (Muller and Grafe,
1978). The lines lacked xanthine dehydrogenase activity but still possessed
the CR activity catalysed by the proximal part of the electron transport
chain of the flavohaemoprotein subunit (Mendel and Muller, 1976, 1979;
pers. comm.). The lines were designated cnx in accordance with the termi-
nology used with A. nidulans mutants of this type (Pateman et al., 1964).
Since it proved impossible to regenerate fertile plants from these four
cnx lines to allow conventional genetic analysis, complementation analysis
of the cnx lines and of the Nia 115 line was carried out by somatic hybrid-
ization (Grafe and Muller, 1983). All nia + cnx combinations resulted in
the formation of cell colonies capable of growing with nitrate as sole
nitrogen source and possessing NR activity. However no nitrate-utilising
colonies were formed from any of the cnx + cnx combinations.
The finding that all nia + cnx combinations studied are complementary
shows that all the mutants involved are recessive. Grafe and Muller (1983)
therefore concluded that the failure of the 4 cnx mutants to complement
each other is not due to dominance but due to allelism. Thus the 4 cnx
mutants tested are allelic to each other, not allelic to Nia 115 (nor to other
nia mutants), and represent four alleles at a gene locus designated cnxA
(Grafe and Muller, 1983). Since these cnx mutants occurred in amphi-
haploid cells and the nia mutants have been shown to carry two unlinked
mutations which affect the duplicate structural genes for the NR apo-
protein, Grafe and Muller (1983) have argued that these cnx mutants are
also double mutants in duplicate genes (cnxA1, cnxA2).
Cnx 135 has now been regenerated to fertile plants. Results of crosses
between Cnx 135 and wild type confirm that Cnx 135 is a double mutant in
a pair of duplicate loci (cnxA1, cnxA2) (Muller, pers. comm.).
Four further NR-minus Mo-co mutants have been isolated from amp hi-
haploid cells of N. tabacum L. var Xanthi by Buchanan and Wray (1982).
Thirty-nine chlorate-resistant lines were isolated but only four lines 042,
P 12, P31 and P47 were unable to grow on nitrate as sole nitrogen source.
These four lines lacked both NR and xanthine dehydrogenase activity. In
vitro complementation between a nia mutant (Nia63) and each of the four
cnx lines (that is formation of NR activity due to assembly of NR mole-
cules when cells of the two NR-minus partners are cohomogenized) and ret-
ention of nitrate inducible CR activity indicates that they possess func-
tional flavohaemoprotein subunits (Buchanan and Wray, 1982; Mendel et
ai., 1984). Genetic analysis has depended on somatic hybrids since it
proved impossible to regenerate plants. All four lines fail to complement
each other but complement an N. tabacum cnxA mutant (Cnx68) as well as
a nia mutant suggesting that they are allelic and recessive (Xuan et ai.,
1983). The gene locus has been designated cnxB.
A total of eight chlorate-resistant NR-minus lines which lack xanthine
dehydrogenase activity have been isolated in N. piumbaginifolia. Plants
could not be regenerated from four lines designated NX (Marton et ai.,
1982 a), and complementation in somatic hybrids showed they belonged to
three complementation groups (NX 1 and NX9; NX21; NX24) (Marton et
ai., 1982 b). The other four lines, which were isolated without a mutagenic
step, also belonged to three complementation groups (CNX20 and
CNX82; CNX27; CNXI03) (Negrutiu et ai., 1983; Dirks et aI., 1985) as
shown by complementation in somatic hybrids. The mutations are
recessive.
b) Complementation Analysis between Nicotiana Molybdenum Cofactor

Mutants
The eight N. tabacum and eight N. piumbaginifolia Mo-co mutants
described above represent a total of four complementation groups, cnxA,
cnxB, cnxC and cnxD, as determined by complementation analysis in
somatic hybrids (Xuan et ai., 1983; Dirks et ai., 1985). The relationship
126 J. L. Wray
between the individual alleles is shown in Table 1. Fertile homozygous

diploid plants have been regenerated from all N. plumbaginifolia CNX
mutants, except CNX 27, and genetic analysis by sexual crosses confirms
their assignment to the complementation groups (Dirks et al., 1985). Indi-
vidual segregation in the F2 generation of the cross between CNX103 and
CNX 20 suggests a linkage between these cnx loci. The nia gene and the cnx
gene responsible for the mutant character of CNX 20 are either located on
different chromosomes or the loci are far enough apart that they behave as
though they are unlinked (Dirks et al., 1985).
c) The Function of Nicotiana cnx Genes

Biochemical analysis of the Nicotiana Mo-co mutants has been carried out
in an attempt to pinpoint the defect in Mo-co synthesis and thus determine
the likely function of these genes.
cnxA alleles of N. tabacum produce a CR species the same size as
wild-type NR, suggesting that these mutants contain a Mo-co which is
defective in catalysis but which is still able to mediate dimerisation of
flavohaemoprotein subunits (Mendel and Muller, 1979). Reconstitution of
NR activity from N. crassa nit-1 extract by Mo-co from N. tabacum and N.
plumbaginifolia cnxA alleles in the presence, but not the absence, of
exogenous molybdate (20 mM) (Mendel et al., 1981 ; Mendel et al., 1986) and
restoration of NR activity to all cnxA alleles tested by growth on medium sup-
plemented with between 0.3 mM and 5 mM molybdate (Mendel et aI., 1981;
Marton et al., 1982a; Negrutiu et al., 1983) indicates that the inactive Mo-co
can be reactivated both in vivo and in vitro by unphysio1ogically high concen-
trations of molybdate. This suggests that the defect in the Mo-co of cnxA
alleles is due to the absence of catalytically active molybdenum within the
cofactor and that the molybdopterin moiety of the Mo-co is probably not
impaired by the mutation. The defect is unlikely to be in molybdenum uptake
since molybdenum levels of cnxA alleles of N. tabacum and N. plumbaginifolia
are similar to wild type (Mendel et al., 1984; Mendel et al., 1986).
Extraction of cells of N. tabacum and N. plumbaginifolia with cnxA
alleles in the presence of reduced glutathione, EDTA and 20 mM
molybdate results in the reconstitution of NR activity (Mendel and Muller,
1985; Mendel et al., 1986). The simplest interpretation of these data is that
the cnxA gene product catalyses the ligan ding of molybdenum to the
molybdopterin moiety of the cofactor. Mendel and Muller (1985) speculate
that the reduced glutathione may supply thiol ligands required for
molybdenum liganding, that EDTA may be involved in molybdenum che-
lation and that, in the presence of these compounds, the enzymatic
insertion step can be bypassed in the mutant. The cnx E mutant of A. nid-
ulans is phenotypically similar to cnxA (Arst et al., 1970).
cnxB alleles of both N. tabacum and N. plumbaginifolia, unlike cnxA
alleles, cannot be repaired in vivo by growth in the presence of unphysio-
logically high levels of molybdate (Buchanan and Wray, 1982; Marton et
al., 1982a; Negrutiu et al., 1983). Since the Mo-co of cnxB alleles of N.
tabacum is capable of in vivo dimerisation of cnxB flavohaemoprotein sub-

units (Buchanan and Wray, 1982) and, in the presence of exogenous
molybdate, is able to reconstitute NR activity from N. crassa nit-l extract,
we originally concluded that the defect in cnxB alleles was not in molyb-
dopterin biosynthesis but in some undefined aspect of molybdenum pro-
cessing (Mendel et at., 1984). However, subsequent studies suggest that
these properties are allele-specific rather than locus-specific since, in con-
trast, the Mo-co of the N. ptumbaginifolia cnxB allele, NX 24, can neither
dimerise its flavohaemoprotein subunits nor reconstitute NR activity from
N. crassa nit-l extract (Mendel et at., 1986). This suggests that the cnxB
mutation lies within the molybdopterin biosynthetic pathway and that
NX 24 represents a null allele at this locus. This is supported by the obser-
vation that fluorescent oxidation products of molybdopterin (Kramer et at.,
1984) cannot be detected in NX24 extracts but are present in extracts of
both N. tabacum cnxB alleles and wild type (Mendel et at., 1986; R.
Mendel, pers. comm.). Two types of allele have also been identified at the
A. nidutans cnxB, cnxF, cnxG and cnxH loci. Dimerisation of flavohaemop-
rotein subunits occurs in only one of the two types (MacDonald et at.,
1974).
The cnxC allele, NX 21, of N. ptumbaginifolia possesses a Mo-co which
is not only defective in catalysis (Marton et at., 1982 a) but also in dimeri-
sation ability, since it cannot dimerise the flavohaemoprotein subunits
present in NX21 nor reconstitute NR activity from N. crassa nit-l extract.
The amounts of fluorescent oxidation products of Mo-co are low, sug-
gesting the defect is in molybdopterin synthesis (Mendel et at., 1986).
The single cnxD allele, CNX 103, of N. ptumbaginifolia cannot be
repaired in vivo by unphysiologically high levels of molybdate (Negrutiu et
at., 1983) but is otherwise uncharacterised.
d) Petunia hybrida
Three Mo-co defective lines representing two putative cnx loci have been
identified by screening mutagenised Petunia cell suspensions for resistance
to chlorate. Two of the lines (3 and 4) are allelic as determined by comple-
mentation in somatic hybrids and are molybdenum-repairable. The other
line (line 2) is non-allelic to lines 3 and 4, and NR activity cannot be re-
stored by growth on unphysiologically high levels of molybdate. All three
lines lack NR and xanthine dehydrogenase activity (Steffen and Schieder,
1984).
e) Hyoscyamus muticus
Unlike all other selection procedures in tissue culture, which have relied on
chlorate screening, mutants in H. muticus have been isolated using a non-
selective total isolation procedure pioneered by Beadle and Tatum (1945)
and subsequently applied successfully to the isolation of mutants in Datura
innoxia (Savage et at., 1979) and the moss Physcomitrella (Ashton and Cove,
1977).
128 J. L. Wray
Four clones, MA2, 12 D 12, VIC2 and XIVE9 were isolated on the basis
of a growth requirement for casein hydrolysate and were subsequently
shown to be both nitrate non-utilizers and resistant to 50 mM chlorate
(Strauss et al., 1981; Gebhardt et al., 1981; Fankhauser et al., 1984). All
four lines are unable to grow on hypoxanthine as sole nitrogen source and
lack xanthine dehydrogenase activity (Fankhauser et al., 1984). Both MA2
and 12 D 12 are molybdate repairable in vivo (Fankhauser et al., 1984).
Recently further Mo-co defective lines (0, Q, I, T and C) have been iso-
lated on the basis of chlorate resistance and they, together with the lines
discussed above, have been analysed in some detail by complementation in
somatic hybrids and by in vitro complementation by co-homogenisation of
mutant lines (H. Fankhauser, pers. comm.). These studies point to the exis-
tence of a total of four complementation groups.
groups which have been designated A (MA-2,

The molybdenum-repairable lines fall into three complementation
and Q), B (I) and C (12
D 12). Mutant I of group B complements neither A nor C. This overlapping
complementation pattern may signify the existence of two mutable loci (A
and C) determining molybdenum ligand binding, with B being a double
mutant. Alternatively all three mutants could be present within one locus
with A and C complementing intragenically. This type of organisation has
not been described in other plants but is similar to the nit-9 ABC locus of
Neurospora (Dunn-Coleman, 1984 a). Lines XIVE9 plus T and C plus
VIC 2 have been assigned to groups D and E respectively and are presently
uncharacterised. Lack of complementation between MA2 and a N.
tabacum cnxA allele, Cnx 68, in interspecific somatic hybrids suggests that
the complementation group A mutation is equivalent to the cnxA mutation
of N. tabacum (Lazar et al., 1983).
f) Hordeum vulgare
Molybdenum-cofactor defective whole plant barley mutants representing
at least three gene loci have been identified. One of these, the nar2 locus, is
represented by the allele nar2a (previously Az 34) which was isolated by
screening individual seedling leaves for lowered in vivo NR activity
(Warner et al., 1977; Kleinhofs et al., 1978). It is a poor source of func-
tional Mo-co in the in vitro reconstitution of NR from a barley apoprotein
source (Narayanan et al., 1984) and since it still possesses low levels of xan-
thine dehydrogenase activity (Somers et al., 1983) and NR activity (8 % of
the wild-type level), it appears to be a leaky Mo-co mutant.
Unlike nar2 a, other Mo-co mutants have been isolated on the basis of
chlorate resistance. Selection for resistance to 8 mM chlorate within an M2
population of barley cv. Winer led to the isolation of two mutants (Xnol8
and Xno19) which had low levels of both NR (Shumny and Tokarev, 1982)
and xanthine dehydrogenase activity (Somers et al., 1983). Like nar2 a,
both Xno18 and Xno19 were poor sources of functional Mo-co (Narayanan
et al., 1984), and genetic analysis by sexual crossing showed them to be
allelic to each other but not to either nar1 (Shumny and Tokarev, 1982) or
nar2 a (Kleinhofs et ai., 1983). Xno 18 and Xno 19 thus represent alleles of a
further Mo-co gene locus which has been designated nar3.
A much more rigorous selection procedure than that used by either
Kleinhofs and co-workers (Warner et ai., 1977; Kleinhofs et ai., 1978) or
Shumny and Tokarev (1982) has led to the recovery of four further Mo-co
mutants, R (= Rothamsted) 9401, 9201,11301 and 12202 (Bright et ai., 1983;
Wray et ai., 1985; B. Steven, pers. comm.). M2 seedlings (mutagenised with
azide in the M t ) were screened for 8 days with 10 mM chlorate and the
selected plants were grown hydroponically with ammonium ions as sole
nitrogen source. Maintenance of the plants to flowering was extremely dif-
ficult and several of the selections were self-infertile. However, three lines
were recovered as heterozygotes after crossing with the cv. Golden Promise,
whilst R 12202 was fertile and was recovered in the homozygous state.
All mutants lacked xanthine dehydrogenase activity and NR activity
(Bright et ai., 1983; B. Steven, unpublished). R9201 and R9401 are allelic
whilst R 11301 represents a separate locus. R 12202 has not yet been tested
(B. Steven, unpublished).
R9201lR9401 are allelic to nar2a (previously Az34) (A. Kleinhofs,
pers. comm.) and thus represent two further alleles at this locus, nar2 band
nar2 c. R9401 lacks dimerised flavohaemoprotein subunits and its Mo-co
is unable to reconstitute NR activity from N. crassa nit-1 extract (Wray et
ai., 1985). The results obtained from R9201 are similar (B. Steven, unpub-
lished). One may draw the tentative conclusion from this that these nar2
alleles possess a defective Mo-co which is unable to efficiently dimerise
flavohaemoprotein subunits, and that the nar2 locus probably specifies a
step in molybdopterin biosynthesis. Evidence from the nar2 a allele, Az 34,
is consistent with this suggestion (Narayanan et ai., 1983; 1984).
The Mo-co of R 11301 is capable of reconstituting NR activity from
nit-l extract in the presence of exogenous molybdate and thus this mutant
is unlikely to be altered in molybdopterin synthesis (B. Steven, unpub-
lished). The biochemical evidence supports the conclusion drawn from
genetic analysis that R 11301 is defective in a different step in Mo-co syn-
thesis from R9401. This step is unlikely to be equivalent to that in N.
tabacum cnxA mutants since neither NR nor xanthine dehydrogenase
activity is restored to R 11301 (nor to R9401 or R9201) by growth on
unphysiologically high levels of molybdate (B. Steven, unpublished).
The R series of mutants discussed above lack NR activity and are con-
ditionallethal mutants. They are able to grow on reduced forms of nitrogen
but show no growth on nitrate as sole nitrogen source and eventually die
(Bright et ai., 1983; B. Steven, unpublished). In contrast, all other barley
mutants discussed above have residual NR activity and grow to different
extents with nitrate as sole nitrogen source.
g) Arabidopsis thaliana
Arabidopsis thaliana was the first plant species in which mutations in
nitrate assimilation were described (Oostindier-Braaksma, 1970; Oostin-
130 J. L. Wray
dier-Braaksma and Feenstra, 1973). These whole plant mutants have

NADH-NR activities ranging from 1-50 % of the wild-type level (Bra-
aksma and Feenstra, 1982 a). Two Mo-co mutants, B-25 (designated rgn)
and B-73 (designated cnx), have been described which have the lowest NR
activities of these mutants and also show the poorest growth on nitrate as
sole nitrogen source. Most of the other mutants grow as well on nitrate as
sole nitrogen source as the wild type.
The xanthine dehydrogenase activity of rgn is about 10-30 % of the
wild type and the CR activity of the flavohaemoprotein subunits sediments
at 4 S, suggesting that the subunits are not dimerised (Braaksma and
Feenstra, 1982 b). This perhaps implies a defect in the molybdopterin
moiety of the Mo-co, although the mutant still retains appreciable levels of
xanthine dehydrogenase activity. cnx has only 10 % of the wild-type xan-
thine dehydrogenase activity but has dimerised flavohaemoprotein sub-
units. Since NR activity and ability to grow on nitrate can be increased by
growth on unphysiologically high levels of molybdate, it is probable that
cnx has a mutation equivalent to that of N. tabacum cnxA and that it is
defective in the insertion of molybdenum into the Mo-co.
Revertants of rgn have been isolated which carry mutations in the same
suppresser gene (su) which is unlinked to rgn. These reverse mutants can
grow on nitrate as sole nitrogen source and are susceptible to chlorate (Bra-
aksma and Feenstra, 1982 c). The revertants have 0.4 to 1.5 times the
wild-type NR activity, and the ability to assemble NR from its subunits,
missing from rgn, has been restored. This appears to be the only case of
suppressor genes so far reported in higher plants, although the nature of
the suppressor mutation is unclear.
h) Pisum sativum
A leaky Mo-co mutant lacking xanthine dehydrogenase activIty but
retaining 20 % of the wild-type level of NADH-NR activity has been iso-
lated by screening an M z population of pea seedlings for lowered in vivo
NR activity (Warner et ai., 1982). The nitrate inducible CR activity and the
nitrite reductase activity of the mutant, A 300, were double the wild-type
induced value probably reflecting the increased level of nitrate present.
The mutation in A300 designated nar2, is monogenic, exhibits incomplete
dominance and is unlinked to the nar 1 locus. The E I mutant has lowered
NR and xanthine dehydrogenase levels and is probably a Mo-co mutant
(Feenstra and Jacobsen, 1980; Jacobsen et ai., 1984).
D. Nitrite Reductase Mutations

No NiR apoprotein gene mutants have been identified so far from higher
plants and it is, of course, unlikely that they could be identified by chlorate
screening since, like wild-type plants, NiR-minus plants possess NR and
would be expected to be sensitive to chlorate. The use of replica plating
techniques has enabled NiR structural gene mutants to be isolated from A.

nidulans (Pateman et al., 1967) and N. crassa (Coddington, 1976) on the
basis of their growth requirements, since such mutants are unable to utilise
either nitrate or nitrite as sole nitrogen source but grow like wild type on
ammonium. Replica plating of this type is difficult with higher plant cells
but other .approaches to the isolation of these mutants do exist. At the
whole plant level one could screen for individuals within M z populations
which accumulate nitrite after the application of nitrate, as has also been
suggested by Dunn-Coleman et al. (1984). The major difficulty here is that
nitrite is toxic, but it might be possible to devise screens based on
short-term exposure of plants to low levels of nitrate which allow NiR-
minus mutants to be recovered. At the callus level a total isolation
procedure might be used. Such l}n approach has already yielded pan-
tothenate auxotrophs in Datura innoxia (Savage et al., 1979), and lines
auxotrophic for histidine, tryptophan, leucine or nicotinamide, as well as
NR-minus nitrate non-utilizers, in Hyoscyamus muticus (Gebhardt et al.,
1981; see also Chapter 10, this Volume).
E. Regulatory Alterations in Nitrate Assimilation

Induction of NR and NiR activity by nitrate in A. nidulans is mediated by
the product of the nirA locus which acts in a positive regulatory fashion
such that it is required for the expression of the niaD and niiA apoprotein
genes and is active only in the presence of nitrate (Pateman et al., 1967;
Cove, 1979). nirA - strains grown in either the presence or absence of
nitrate produce low levels of both NR and NiR equivalent to those of the
wild-type strain grown in the absence of nitrate. NirAc strains have consti-
tutive synthesis of both NR and NiR. Repression by ammonium or glut-
amine (nitrogen metabolite repression) is mediated by the areA gene
product (Arst and Cove, 1973). A locus performing the same or similar
function has been identified also in N. crassa (Tomsett and Garrett, 1980;
Tomsett et al., 1981).
Nitrate reducta~e regulatory gene mutants have not been isolated, or at
least have not been recognised, in higher plants so far. The only possible
exception to this generalisation may be the nr, mutants of soybean (Nelson
et al., 1984). There is, however, some evidence that mutations within the
apoprotein gene, or within Mo-co genes, lead to regulatory perturbations
which may have their parallels in both A. nidulans and N. crassa.
In two cnxA alleles of N. tabacum Cnx6812 (a subline of Cnx68), and
Cnx 101 (Mendel and Miiller, 1979), NR apoprotein is produced consti-
tutively, as measured by the presence and amount of a 7.7 S CR species in
nitrate-less cell culture and, in the case of Cnx68/2 and Cnx 109, NiR
activity is also produced constitutively. NiR levels were still inducible in
Cnx 101 (and also in Nia 102) but the activity present, both under unin-
duced and induced conditions, was double that of the wild-type values.
NR-associated CR activity was also produced constitutively in cnxB alleles
(Mendel et al., 1984). Whether the constitutive levels of NR-associated CR
132 J. L. Wray
and of NiR activity are a consequence of the Mo-co mutation (or the nia
mutation in the case of Nia 102) is not clear. However in A. nidulans some
mutations causing loss of a functional NR molecule lead to constitutive
synthesis of both NR and NiR and suggest an autoregulatory role for func-
tional NR molecules in the regulation of synthesis of both NR and NiR
(Pateman et al., 1967; Cove and Pateman, 1969).
Cove (1979) has proposed a model to accommodate these findings.
According to this model, in the absence of nitrate NR interacts with the
nirA gene product converting it into a form which does not allow the niaD
and niiA genes to be expressed. However, in the presence of nitrate NR
binds to its substrate and can no longer interact with tlie nirA gene product,
thus allowing it to mediate expression of the niaD and niiA genes. Many
mutants possessing defective NR molecules, due to mutation either in the
apoprotein gene or a Mo-co gene, would no longer be able to inactivate the
nirA gene product and would show constitutive synthesis of both NR and
NiR. Those mutants which possess wild-type regulation have catalytically
defective NR molecules which can still interact with the nirA gene product
in the wild-type manner.
Whilst the data from N. tabacum provides some evidence for an auto-
regulatory role for NR, there is no a priori reason to suppose that the regu-
latory networks involved in controlling NR and NiR gene expression in
higher plants are the same as those operating in lower eukaryotes and there
is in fact little corroborative evidence from other sources. Thus NR-asso-
ciated CR and/or NiR levels are not constitutive in nar 1 and nar2 alleles
of barley cv. Steptoe (KJeinhofs et al., 1980), in the Mo-co mutant R9401 of
barley cv. Golden Promise (Bright et al., 1983), in the nr1 mutant of soybean
(Nelson et al., 1984) nor in the narl and nar2 mutants of pea (Warner et
al., 1982). However NR-associated CR and/or NiR levels have not been
assayed from uninduced tissue of the large number of other mutants so far
isolated.
There is evidence to suggest a link between the synthesis of functional
NR molecules and of the Mo-co. The Mo-co of N. tabacum (Mendel et al.,
1982) and of barley (Narayanan et al., 1984; Mendel et al., 1985) is nitrate
inducible. Some N. tabacum nia alleles, however, have a Mo-co which is no
longer nitrate-inducible whilst others, which retain residual benzyl vio-
logen-NR activity, still possess a nitrate-inducible Mo-co (Mendel et al.,
1982). The authors have speculated that regulation is normal in the latter
group due to the formation of an assembled, but catalytically-defective,
NR holoenzyme which is still able to bring about wild-type regulation,
whilst the defect in the apoprotein of the former class does not allow
assembly and wild-type regulation is lost. This hypothesis suggests that the
NR holoenzyme acts in a positive way to regulate Mo-co synthesis. This is
supported by the observation that Mo-co synthesis in cnxA (Mendel et al.,
1982) and cnxB (Mendel et al., 1984) alleles of N. tabacum which produce
an inactive NR holoenzyme constitutively is also constitutive. The possi-
bility that this phenomenon exists in other species appears not to have
been examined yet.
F. Conclusions
A great number of mutants altered in nitrate reduction (due either to nia or

cnx mutants) have been isolated usually on the basis of chlorate resistance
(Table 1). Selection pressure appears to playa role in the type of mutant
obtained at the whole plant level. Rigorous screening of M2 barley plants
by Bright et al. (1983) (10 mM chlorate for nine days) led to the recovery of
only Mo-co mutants (B. Steven, unpublished), whilst the least rigorous
selection procedure (direct selection of barley plants showing decreased
levels of NR (Kleinhofs et al., 1980) recovered many more apoprotein gene
mutants than Mo-co mutants.
At least part of the reason may be due to the fact that barley (or at least
the cultivar Steptoe) possesses a second NR apoprotein gene (encoding the
NADPH-linked enzyme) which is expressed only in the presence of a
defect in the NADH-NR apoprotein gene (Dailey et al., 1982a, b).
NADH-NR apoprotein gene mutants possessing a functional NADPH-NR
apoprotein gene are still chlorate sensitive (Warner and Kleinhofs, 1981).
Whilst double apoprotein gene mutants would be expected to be chlorate
resistant, their presence in M2 populations is likely to be extremely low.
Such mutants isolated at the cell culture level in N. tabacum occur at a fre-
quency of 10- 7 (Muller, 1983). This argument suggests that the Mo-co
mutants isolated by Bright et al. (1983) are the result of single gene muta-
tions.
This does not explain, however, why so few Mo-co mutants were re-
covered by Kleinhofs et al. (1980) nor why they are often poorly repre-
sented when selection is done at the cell/protoplast level (Muller and
Grafe, 1978; Marton et al., 1982a; Negrutiu et al., 1983) since there are
likely to be at least six Mo-co gene loci compared to only one or two struc-
tural gene loci. One possible explanation is that nia loci are more prone to
mutagenesis than are cnx loci. Alternatively it may be related to the under-
lying mechanism involved in chlorate resistance. Cove (1976b) has shown
that in Aspergillus the ratio of nia to cnx mutants is influenced by the
nitrogen source on which selection is made. It is clear, however, that a
number of different Mo-co gene mutants have been isolated in higher
plants on the basis of chlorate resistance and it is likely that all the cnx loci
described in Aspergillus will eventually be shown to have their counterpart
in the plant genome.
Nitrate uptake (Doddema and Telkamp, 1979) and regulatory gene
mutations (Ryan et al., 1983; Nelson et al., 1984) are poorly represented, if
at all, amongst the nitrate assimilation mutants so far identified. It is unfor-
tunate that the large number of chlorate resistant clones which retain NR
activity (Muller and Grafe, 1978; Murphy and Imbrie, 1981; Marton et al.,
1982a; Buchanan and Wray, 1982) have not been examined since, as dis-
cussed above, it could be that some of these lines carry either regulatory or
nitrate uptake mutations. Buchanan and Wray (1982) have isolated a cell
line, 041, on the basis of chlorate resistance which retains NR activity.
Growth. kinetics on nitrate as sole nitrogen source are different from the
134 J. L. Wray
wild type as are nitrate and NR levels, although growth of 041 and
wild-type cells on either glutamine or ammonium is the same (Qureshi et
al., 1982; J. A. Qureshi and J. L. Wray, unpublished). It is clear that nitrate
assimilation is altered in the 041 cell line although whether this is due to
uptake or regulatory perturbations is uncertain.
With only a few exceptions mutations have been induced with base sub-
stitution mutagens. It may be that other types of mutagenesis would yield
other types of mutations, such as regulatory mutations. Insertion mutag-
enesis is a possible technique which might be carried out either with plant
controlling elements or even with the T-DNA region of the Agrobacterium
tumefaciens Ti plasmid. Deletion mutants might be present amongst NR-
minus progeny regenerated from tissue culture (Larkin and Scowcroft,
1981).
One disadvantage of selecting mutants in cell culture is that it may not
be possible to regenerate whole, fertile plants. This makes it impossible to
study the mutation by conventional genetic analysis. This is unfortunate
since it has been argued that the most satisfactory evidence that a muta-
tional event has in fact occurred relies on the regeneration of fertile plants
from the cell lines and transmission of the variant trait in sexual crosses
and that until this has been carried out the isolates should be described as
"phenotypic variants" or "cell lines" rather than as mutants (Maliga, 1976).
Inability to regenerate mutants also precludes the possibility of studying
the effect of the mutation on the physiology and biochemistry of the intact
plant and limits their usefulness in somatic hybridization and plant trans-
formation.
The evidence summarized in Table 1 suggests that it might be more dif-
ficult to regenerate cnx than nia mutant cell lines to fertile plants but it is at
present not clear whether this is directly related to the cnx genotype.
IV. Applied Aspects
A. Somatic Hybridisation
Glimelius et al. (1978) were the first to show that somatic hybrids resulting
from PEG fusion of protoplasts derived from N. tabacum L. var. Gaters-
leben NR-minus nia and cnx cell lines could be recovered on the basis of
their growth on medium containing nitrate as sole nitrogen source. The
possibilities of reversion or of cross-feeding were excluded as possible
explanations for the occurrence of nitrate-utilising colonies. In recon-
struction experiments one wild-type colony amongst 4 x 10- 4 nia colonies
could be selected on nitrate medium at nearly 100 % efficiency (Pental et
al., 1982). Reversion of Nia30 occurs at rates of about 10- 6 CR. Grafe, pers.
comm.). These results suggest that the NR-minus phenotype might be
useful as a selection marker in the identification of somatic hybrids and
perhaps also in plant cell transformation studies.
Complementation in somatic hybrids via protoplast fusion has been
very useful in determining allelism of NR mutant cell lines, and intraspe-

cific, interspecific and intergeneric complementation has been reported.
Biasini and Marton (1985) have described a rapid assay for genetic comple-
mentation of NR deficiency via bulk protoplast fusion. Several workers
have reported the use of cnx mutant protoplasts as one partner in somatic
hybrid selection schemes (Wallin et aI., 1979; Glimelius et al., 1981; Gli-
melius and Bonnett, 1981; Hein et al., 1983). Hamill et al. (1983) have con-
structed an NR-minus streptomycin resistant double mutant of N. tabacum.
On selective medium containing nitrate as sole nitrogen source and also
streptomycin it is likely that only nuclear/cytoplasmic recombinants could
grow as green colonies when the NR-minus, SR-plus protoplasts are fused
with wild-type protoplasts of other species. Thus the double mutant might
be useful in constructing new nuclear/cytoplasmic gene combinations.
Although it was hoped that somatic hybridization by protoplast fusion
would lead to unique hybrid plants impossible to obtain by conventional
sexual hybridization, limited success has been achieved only with closely
related species (Schieder and Vasil, 1980). This is due largely to somatic
incompatibility problems between distantly related or unrelated genomes
and is manifested in the failure of hybrid development and in the elimi-
nation of chromosomes from both cell hybrids and regenerants. Attempts
have therefore been made to bypass somatic incompatibility barriers
between distantly related plants by the transfer of limited amounts of
genetic information. As Shephard et al. (1983) have argued, introgression
of genes from diverse alien species could significantly expand germplasm
pools for such characters as pest or stress resistance provided that the
introduced genes were expressed and were capable of being manipulated
by breeding techniques.
Gupta et al. (1982) have described the transfer of nuclear genes from
Physalis minima and Datura innoxia to N. tabacum. When X-irradiated,
mitotically inactive protoplasts of P. minima or D. innoxia were fused with
Cnx 68 mutant protoplasts, cell lines were isolated able to grow on nitrate
as sole nitrogen source. Restoration of NR and xanthine dehydrogenase
activity to the Cnx 68 mutant is probably due to transfer and expression of
the equivalent genes from the X-irradiated fusion partner. Whilst this
procedure might have application for introducing desirable characters
from diverse genomes, it is clearly a relatively crude method of gene
transfer which does not allow the investigator total control over the genes
being transferred. Much more precise procedures for gene transfer exist
and some of these are briefly discussed below.
B. Cloning of Nitrate Assimilation Genes

Molecular cloning of a higher plant NR apoprotein gene or of a Mo-co gene
has not yet been accomplished although there is intense activity in this area.
The approaches which might be used to clone Mo-co genes are more limited
than those for apoprotein gene cloning since we have no information on any
higher plant Mo-co gene-product. This means that techniques for the iso-
l36 1. L. Wray
lation of genes which depend either on the use of antibody probes or of syn-
thetic oligonucleotide probes cannot be employed.
Several general approaches to either apoprotein or Mo-co gene cloning
are available. One possibility is the use of "gene-tagging". This has been
used to clone an NR gene from the non-nitrogen fixing cyanobacterium
Anaeystis nidulans. Eight mutants representing three loci designated narA,
narB and nmC were isolated by insertion mutagenesis of A. nidulans with
the transposon Tn901, which encodes ampicillin resistance (Kuhlmeier et
al., 1984). Whether these are apoprotein gene or Mo-co gene loci is
unknown but the mutants could not be repaired by growth on medium sup-
plemented with I mM molybdate. The narB gene was cloned in the fol-
lowing way. The DNA of a narB mutant allele, Nar6, was restricted with
EeoRI which does not cleave in the transposon sequence. The EeoRI frag-
ments were then cloned into the E. coli vector pACYC 184 and, after trans-
formation, selection was made for ampicillin-resistant colonies. The
plasmid from one such colony, designated pNRT63, was able to transform
the Nar6 allele to the wild-type phenotype and contained a 20 kb EeoRI
insert. The transposon was located on a 9.2 kb Sail fragment. pNRT63
was then used as a probe against a wild-type genomic cosmid library.
Several cosmids were isolated which were able to transform the Nar6
mutant to the wild-type phenotype. One of these, pNR63, was found to
contain a Sail fragment with the expected size of 5.1 kb, i. e. the size of the
pNRT631 Sail fragment minus the size of the transposon. Cloning of the
5.1 kb Sail fragment into pACYC 184 gave pNR631, which also trans-
formed the Nar6 allele to the wild-type phenotype. The insert of pNR631
hybridised to a unique chromosomal fragment, suggesting that the nar B
gene is represented by a single-copy.
A possible candidate "tag" for higher plants is the controlling element
Mu I found only in Robertson's Mutator maize lines (Robertson, 1978;
Bennetzen, 1984). Mu 1 produces mutations at most loci investigated at
rates near 10-4, 30-50 fold higher than in non-Mutator lines (Robertson,
1978). Alcohol dehydrogenase mutants have been induced by insertion of
Mu 1 into the adh gene (Strommer et al., 1982; Freeling et al., 1982) and it is
not unreasonable to suppose that NR-minus mutants could also be gen-
erated. DNA isolated from such mutants could be probed with the cloned
Mu 1 sequence and plant genomic sequences carrying a Mu 1 insertion iso-
lated. However, since Mutator lines carry 15 -40 copies of Mu 1 per diploid
genome (Bennetzen, 1984) Mu 1 will be inserted in numerous other places
besides the NR structural gene or a Mo-co gene. This would considerably
complicate the identification of the genomic NR-related sequence carrying
a Mu 1 insertion. Further complication is introduced by the fact that Mu 1
insertions are unstable. Recently, however, the a I locus of Zea mays has
been cloned using Mu 1 and other controlling elements, Spm-18 and En 1,
as gene tags (O'Reilly et al., 1985). "Tagging" with the T-DNA of the Ti
plasmid of Agrobaeterium tumefaciens might be useful in Nieotiana species
and other dicotyledonous plants.
A further general approach is by functional complementation of plant
NR-minus mutants but this requires an appropriate system for the transfer
and expression of plant DNA sequences and for the identification of trans-
formants. Since most of the plant NR-minus mutants are in Nicotiana
species, a system based on the Ti plasmid of Agrobacterium tumefaciens
might be suitable (reviewed in Schell and van Montagu, 1983). A plant
genomic library might be constructed in T-DNA from which oncogenic
functions have been deleted (Zambryski et al., 1983) and which carries an
antibiotic resistance marker (Herrera-Estrella et al., 1983; Fraley et al.,
1983). Initial screening for transformants would be on the basis of anti-
biotic resistance followed by identification of those transformants which
were able to utilize nitrate as sole nitrogen source. Due to the nature of this
system, complementing DNA sequences would be integrated into the host
chromosome and appropriate procedures would be needed to rescue them.
Over the past three years transformation systems have been developed
for Aspergillus nidulans (Ballance et al., 1983; Tilburn et al., 1984; John-
stone et al., 1985) which make use of a cloned gene on appropriate vectors
to functionally complement the corresponding mutant A. nidulans recipient
to the wild-type phenotype. Transformation is by integration but spon-
taneous excision of integrated plasmids is frequent enough to allow their
re-isolation in E. coli and has permitted the molecular cloning of the A. nid-
ulans developmental gene, brlA (Johnstone et al., 1985). As indicated
above, a wealth of nitrate assimilation mutants is available in A. nidulans
and it is likely that one or other of these transformation systems will
shortly allow the cloning of A. nidulans nitrate assimilation genes.
Cloning of Mo-co genes from E. coli and from N. crassa has in fact
already been accomplished by functional complementation of equivalent
E. coli Mo-co mutants. The Mo-co is highly conserved between eukaryotes
and prokaryotes and the chlA, chlB, chiD, chlE and chlG loci have been
implicated in the synthesis of the E. coli Mo-co (Amy, 1981). Giordano et
al. (1981) first cloned the E. coli chlA gene and, more recently, Kleinhofs et
al. (1983) and Taylor et al. (1983) have shown that the chlA locus is
divisible into two components, one of which was designated chlM (and
which could complement the chlA mutant, SA493) and chIN (which could
complement the chlA mutant, JP382). A 1.9 kb Bell fragment in pBR322
complemented the mutant SA493 when inserted in either orientation sug-
gesting that the entire gene was present and was active from its own
promoter. In vivo transcription-translation of the 1.9 kb Bell fragment in
maxicells and minicells identified the chlM gene product as a ca. 15,000
mol. wt. polypeptide (Kleinhofs et al., 1983; Taylor et al., 1983).
Kleinhofs et al. (1983) chose to clone the chlA gene since they believed
it was equivalent to the mutated gene in the N. tabacum cnxA allele,
Cnx68, (Muller and Grafe, 1978) and the Hyoscyamus muticus mutant
MA-2 (Strauss et al., 1981). Extracts from E. coli wild type and chlB, chIC,
chlE and chlG, but not chlA, mutants were able to reconstitute NR activity
in each of the plant mutant extracts. Dunn-Coleman (1984b) has recently
argued that the chID locus rather than the chlA locus is equivalent in
function to the N. tabacum cnxA locus since E. coli chID mutants, like N.
138 J. L. Wray
tabacum chlA mutants, can be repaired by growth on unphysiologically

high levels of molybdate (Glaser and DeMoss, 1971).
Several fungal genes have been cloned on the basis of their ability to
functionally complement equivalent E. coli mutations. From N. crassa these
include the qa2 gene encoding catabolic dehydrogenase (Vapnek et al.,
1977), the pyr4 gene encoding orotidine 5f-phosphate carboxylase (Buxton
and Radford, 1983), and the trifunctional trp 1 gene (Schechtman and
Yanofsky, 1983). Kinghorn and Hawkins (1982) have cloned the synthetic
dehydrogenase function of the A. nidulans arom cluster. Expression of
these genes in E. coli is likely to be due to fortuitous recognition of accom-
panying fungal promoters. It is unlikely that this approach will be of use as
a general procedure for fungal gene cloning.
However, Dunn-Coleman (1984b) has taken a similar approach to
clone a N. crassa Mo-co gene. He constructed an N. crassa genomic library
in pRK9, a derivative of pBR332, and used it to transform two E. coli chiD
Mo-co mutations (chiD and chID::Mu) to wild type. Three independent
transformants examined had plasmids carrying an identical 4.2 kb insert.
Southern blot analysis with one of the transforming plasmids, pMoCo, 1 :4,
showed hybridization to a single band of N. crassa genomic DNA. When
pMo-co, 1:4 was used to transform various E. coli nitrate reductase mutants
(chIA, chlB, chiC, chiD, chlE, chlG and chlM) the pMo-Co plasmid was
capable of restoring NR activity to only the chiD mutant. Mo-co from
chiD; pMo-co cell free extracts was capable of reconstituting more
NADPH-NR activity from extracts of the N. crassa mutants nit-I, nit-7 and
nit-8 than was cell-free extract from the original E. coli chiD mutant. The
chiD mutant of E. coli can be repaired by growth on unphysiologically high
levels of molybdate like mutants of the nit-9 locus of N. crassa (nit-9 A, nit-
9 Band nit-9 C) (Dunn-Coleman, 1984 a). This suggests that the 4.9 kb
insert of pMp-co, 1 :4 might carry sequences equivalent to the nit-9 locus of
N. crassa.
Other approaches which might be used for apoprotein gene cloning exist.
Differential screening of a cDNA library prepared from polyA + RNA
enriched for NR mRNA might be employed. Colonies showing a stronger
hybridization signal with a 'nitrate-induced' cDNA probe than an 'unin-
duced' cD NA probe would be selected as putative NR cD N A clones. Identity
might be confirmed by hybrid-select-translation (Maniatis et al., 1982) and
immune precipitation with NR antiserum. A problem with this approach is
that the abundance of the NRmRNAislikelyto be very low (0.005-0.01 %of
the total poly A + RNA) and its large size (greater than 3 kb) suggests it will be
unstable. It is, however, encouraging that molecular cloning of oat phy-
tochrome cDNA clones has recently been accomplished using an equivalent
approach (Hershey et al., 1984) since its apoprotein is approximately the
same size as that of NR (Vierstra and Quail, 1983) and its mRNA constitutes
only 0.005 % of the total polyA + RNA (Gottman and Schafer, 1983). Hurlburt
and Garrett (1985) have recently isolated eight nitrate-inducible DNA
sequences in Aspergillus using this differential approach.
Two techniques have recently been applied to the identification of
cDNA clones derived from mRNA species present in low abundance. One
of these, which may be used if mono specific antiserum to NR is available,
is the immunological screening of expression libraries constructed in either
plasmid, for example pUC 8 (Helfman et al., 1983), or phage cloning
vehicles. Several workers have in fact reported the production of poly-
clonal antisera to higher plant NR (Graf et al., 1975; Smarrelli and
Campbell, 1981; Kuo et al., 1981) whilst Notton et al., (1985) have pre-
pared monoclonal antibodies to the spinach enzyme.
An example of a phage cloning vehicle is the lambdoid expression
vector, A gtll (Young and Davies, 1983 a). This vector permits the insertion
of cDNA or genomic DNA into a unique EeoR 1 cleavage site located
within laeZ, 53 bp upstream of the p-galactosidase termination codon.
Phage containing inserts can be induced to produce an inactive p-galacto-
sidase protein fused to antigenic protein specified by the foreign DNA,
with the lac inducer IPTG. Screening with specific antibody probes allows
the detection of antigen in populations of up to 106 lysogens per 82 mm
nitrocellulose filter and thus this system should be ideal for the identifi-
cation of NR cDNA clones. The yeast RNA polymerase II gene (Young
and Davies, 1983 b), human p-glucocerebrosidase gene (Ginns et al., 1984)
and the gene encoding the y-subunit of bovine retinal GTPase (Yatsunami
et al., 1985) have been cloned using this approach.
The other technique requires a knowledge of the partial or complete
amino acid sequence of the NR protein. A mixture of synthetic oligonucle-
otides which represents all possible coding combinations for a small
portion of the amino acid sequence can be used directly as a probe for the
detection of unique genes in Southern blot filter hybridization and in
colony and bacteriophage library screening (Wallace et al., 1981; Suggs et
a!., 1981). Alternatively the oligonucleotides can be used indirectly as
primers for cDNA synthesis (Sood et al., 1981).
The abundance of NiR mRNA is likely to be of the same order as that
of the NR mRNA, suggesting that immunological screening of expression
libraries or the use of synthetic oligonucleotide probes might be the best
approach for cloning the plant NiR apoprotein gene. NiR has been
purified from several plant species (Hucklesby et al., 1976; Gupta and
Beevers, 1984; Small and Gray, 1984) and antibodies produced. Cloning
strategies for plant nitrate uptake genes are very limited since the protein
products of the genes have not been identified and cloning by functional
complementation is confounded by the very few and poorly characterised
nitrate uptake mutants available in higher plants and Aspergillus.
An approach which may be utilised here, and which may also be appli-
cable to NR and NiR, is the use of equivalent cloned genes from other
organisms as heterologous DNA probes. Clearly for this approach to be
successful there has to be considerable nucleotide homology between the
DNA sequences involved. Functional- or amino acid sequence-homology
of the gene products is not sufficient. Shah et al. (1983) have isolated actin
genes from genomic libraries of two highly divergent plants, maize and
soybean, using a Dictyostelium actin gene as a heterologous probe. The
140 J. L. Wray
nitrate assimilation genes of Aspergillus are linked and ordered in the

sequence crnA niiA niaD (Tomsett and Cove, 1979; Brownlee and Arst,
1983). Some genomic sequences which functionally complement Aspergillus
niaD or niiA mutants might be expected to carry the crnA gene. If not it
would be relatively simple to fish-out the crnA sequence from an Asper-
gillus genomic library using either the cloned niiA or niaD gene as a probe.
Cloned nitrate assimilation genes will allow a path to be cut through
the jungle of physiological data and permit studies on gene structure, orga-
nisation, and regulation in response to environmental parameters. The pos-
sibility that the heterologous DNA probe approach may permit the ident-
ification of a plant gene equivalent to the Aspergillus regulatory gene, nir A,
is particularly exciting.
C. Plant Gene Transfer Systems

Gene transfer systems based on the Ti plasmid of Agrobacterium tumefa-
ciens, from which oncogenic functions have been deleted, have been
described recently by several groups. These systems allow the transfor-
mation of either protoplasts (De Block et al., 1984; An et al., 1985) or leaf
disc cells (Horsch et al., 1985). Transformants are identified on the basis of
antibiotic resistance markers carried on the T-region. Regenerated plants
express these resistance genes and transmit them to their progeny in a Men-
delian fashion. Vectors of this type allow the transformation of plant
species which can be infected by Agrobacterium (tobacco, petunia and
tomato in the cases above).
A novel gene transfer vehicle based on an E. coli plasmid and con-
taining an antibotic resistance gene marker under the control of the CMV
gene VI expression signal has recently been described (Paszowski et al.,
1984). This allows Agrobacterium-independent direct gene transfer into
plants and transmission of the antibiotic gene to the sexual offspring in a
Mendelian fashion. Systems of these types should allow the transfer of NR
structural genes and Mo-co genes to plants. One might also envisage NR
apoprotein genes or Mo-co genes, instead of antibiotic resistance genes, as
selectable markers in vector-mediated gene transfer to either nia- or
cnx-type recipients. Transformants would regain the ability to utilise nitrate
as sole nitrogen source and could be selected as nitrate-utilizers.
The Ti plasmid system has been extremely useful in the study of regu-
lated plant genes and in promoter mapping. A bacterial chloramphenicol
acetyltransferase gene has been linked to the 5' portion of Pisum sativum
ribulose 1,5-bisphosphate carboxylase small subunit gene and introduced
into N. tabacum via the Ti system. The chimaeric gene is light inducible
(Herrera-Estrella et al., 1984). Introduction of derivatives of the gene into
Petunia cells indicated that a conserved 33-base pair sequence close to the
TATA box of the gene is sufficient to confer light inducibility (Morelli et
al., 1985). Studies of this type should allow the identification of nitrate
regulatory sequences when cloned NR and NiR apoprotein genes become
available.
D. Whole Plant Studies
Studies at the physiological and biochemical level show that nitrate uptake
and nitrate flux, as well as NR activity, correlate to some extent with accu-
mulation of reduced nitrogen by the plant (Reed and Hageman, 1980 a, b;
Shaner and Boyer, 1976). Partitioning of nitrate between storage and meta-
bolic pools (Ferrari et al., 1973) and between different plant parts, and the
ability of the plant to mobilise stored nitrate and redistribute reduced
nitrogen to the grain (Dalling et al., 1975) must also be related in some way
to accumulation of grain reduced nitrogen. However, since the ability of
the plant to acquire nitrate from the environment influences both nitrate
flux and also NR activity, it is clear that nitrate uptake is a major point of
control in the assimilatory process.
Estimates show that more than two-thirds of the edible dry matter and
half the protein produced in the world are contributed by the cereals
(Evans, 1975). The increased use of fertiliser nitrogen has been a significant
factor in increasing productivity of these crops. However, concerns about
the high cost of energy required to produce fertiliser nitrogen and the effect
of nitrate pollution on the quality of the environment (Foster et al., 1982;
Magee, 1982; Wilkinson and Greene, 1982) calls for more efficient
management and utilization of fertiliser nitrogen. These objectives can be
approached at the plant level by improving the efficiency of nitrate utili-
zation through an understanding of the molecular processes involved in the
acquisition and assimilation of nitrate nitrogen and the factors influencing
these processes in the plant. Our present inability to analyse nitrate uptake
in molecular terms is particularly disappointing.
Grain yield, protein yield, grain reduced nitrogen or plant reduced
nitrogen have been shown to be related to NR levels in a wide variety of
crop-plants and some workers, for example Johnson et al. (1976), have sug-
gested that NR levels might be useful as a predictive indicator of grain
yield. However, the correlation values are often low and grain yield and
plant reduced nitrogen were little affected in NR-deficient mutants of
barley (Oh et al., 1980; Warner and Kleinhofs, 1981) possessing less than
10 % of wild-type NR activity. The ability to introduce multiple copies of
cloned NR apoprotein genes into crop plants would allow the effect of NR
gene dosage to be analyzed in the same genetic background and perhaps
finally resolve this question. Of relevance here is the observation of Muller
(1983) that in tobacco the level of NR activity is independent of the
number of nia+ genes, indicating complete compensation of gene dosage
effects by regulatory mechanisms.
An alternate approach might be to introduce NR apoprotein genes
which have more efficient promoters or which are kinetically more efficient
either as the result of natural variation or due to site-directed mutagenesis.
In this respect the NR of Erythrina senegalensis is of some interest. Nitrate
reductase levels in this tropical leguminous tree are some 100-fold those of
most crop plants (Stewart and Orebamjo, 1979). Characterisation of the
cloned gene would allow the reasons for this to be determined.
142 J. L. Wray
Field-grown soybeans reportedly obtain only 25 % of their nitrogen

from atmospheric dinitrogen, the remainder being derived from soil
reserves of nitrate and ammonium (Hardy et al., 1968). Nitrate retards nod-
ulation and nodule development (Munns, 1968), causes nodule senescence
(Chen and Phillips, 1977) and reduces dinitrogen fixation. Understanding
how nitrate brings about these effects is important for devising strategies
for maximum dinitrogen fixation. The responses could be due to nitrate
itself or due to the products of the reduction of nitrate by the plant or
bacteroid NR. No decrease in inhibitory effects of nitrate were seen when
legumes were inoculated with NR deficient mutants of Rhizobium sug-
gesting that bacteroid NR was not involved (Gibson and Pagan, 1977;
Manhart and Wong, 1980). Attempts to distinguish between the inhibitory
effects of nitrate and of its reduction products by treating soybeans with
tungstate were complicated by the fact that both NR activity and nitrate
uptake were inhibited (Harper and Nicholas, 1978).
Feenstra et al. (1982) have used the El mutant of pea, which possesses
20 % of the wild-type activity but has normal nitrate uptake, in an attempt
to distinguish between the two possibilities. Whilst ammonium ions inhib-
ited acetylene reduction to the same extent in both the wild type and the
E 1 mutant the inhibitory effect of nitrate was much reduced suggesting
that the effects are due in part to the product(s) of nitrate reduction. This is
unlikely to be nitrite (Streeter, 1985).
More recently, Jacobsen and Feenstra (1984) have isolated a mutant
from M z populations of pea seedlings on the basis of its ability, unlike
wild-type plants, to nodulate abundantly in the presence of 15 mM nitrate.
This line is unlike other 'highly nodulating' lines carrying the nodi and nodz
genes which produce abundant nodules only in the absence of nitrate. The
mutant character is monogenic and recessive and the new gene has been
designated nod3 The biochemical basis of the mutation is unknown. The
results of studies with a pea line carrying both nodi and nodz shows that
nodule mass and seed yield are positively correlated and suggests that the
mutant character of nod3 may be of economic significance.
The ureides allantoin and allantoic acid are important transport forms
of nitrogen in some legumes such as Phaseolus vulgaris and soybean where
they can make up to 80 % of the total sap nitrogen in nodulated plants
(McClure and Israel, 1979). Synthesis of ureides within the nodules is
believed to be via purine breakdown and involves the xanthine dehy-
drogenase-catalysed conversion of xanthine to uric acid (Triplett, 1985)
and its further metabolism to allantoin and allantoic acid by uricase and
allantoinase respectively (Thomas and Schrader, 1981; Reynolds et al.,
1982). Use of Mo-co defective mutants of soybean would be invaluable in
confirming this proposed pathway and in determining whether nodule xan-
thine dehydrogenase is a plant or Rhizobium encoded function. If the
former, then these mutants should provide a much better tool than the xan-
thine dehydrogenase inhibitor, allopurinol, (Fujihara and Yamaguchi,
1978) for studying the long-term role of ureides in the nitrogen economy of
the plant and the adaptation in nitrogen metabolism and transport which
might occur as a result of the inability to synthesise them. However, Mo-co

mutants have not so far been reported from ureide-transporting legumes.
(Chapter 6 in this volume deals further with genetic approaches to nodu-
lation and nitrogen fixation in legumes.)
Acknowledgements
I thank all those colleagues who were prepared to provide me with un-
published details of their work and, further, Dr. C. Deane-Drummond, Dr.
R.-R. Mendel and Dr. A. J. Muller for useful discussions. This review was
written whilst I was in receipt of research grant AG 49/23 from the United
Kingdom Agricultural and Food Research Council. I am indebted to Mar-
garet Wilson for typing the manuscript.
v. References
Aberg, B., 1974: On the mechanism of the toxic action of chlorates and some
related substances upon young wheat plants. Kungl. Lantbrukshogkolans Ann.
15,37-107.
Amy, N. K., 1981: Identification of the molybdenum cofactor in chlorate-resistant
mutants of Escherichia coli. 1. Bacteriol. 148,274-282.
Amy, N. K., Rajagopalan, K. V., 1979: Characterization of molybdenum cofactor
from Escherichia coli. 1. Bacteriol. 140, 114-124.
An, G., Watson, B. D., Stachel, S., Gordon, M. P., Nester, E. W., 1985: New cloning
vehicles for transformation of higher plants. EMBO 1. 4, 277 -284.
Arst, H. N., Cove, D. 1., 1973: Nitrogen metabolite repression in Aspergillus nid-
ulans. Mol. Gen. Genet. 126, 111-141.
Arst, H. N., MacDonald, D. W., Cove, D. 1., 1970: Molybdate metabolism in Asper-
gillus nidulans. I. Mutations affecting nitrate reductase and/or xanthine dehy-
drogenase. Mol. Gen. Genet. 108, 129-145.
Arst, H. N., Tollervey, D. W, Sealy-Lewis, H. M., 1982: A possible regulatory gene
for the molybdenum-containing cofactor in A. nidulans. 1. Gen. Microbiol. 128,
1083-1093.
Ashton, N. W., Cove, D. 1., 1977: The isolation and preliminary characterisation of
auxotrophic and analogue resistant mutants of the moss, Physcomitrella patens.
Mol. Gen. Genet. 154, 87 -95.
Ballance, D. 1., Buxton, F. P., Turner, G., 1983: Transformation of Aspergillus nid-
ulans by the orotidine-5'-phosphate decarboxylase gene of Neurospora crassa.
Biochem. Biophys. Res. Commun. 112, 284-289.
Beadle, G. W., Tatum, E. L., 1945: Neurospora II. Methods of producing and
detecting mutations concerned with nutritional requirements. Am. 1. Bot. 32,
678-686.
Beevers, L., Hageman, R. H., 1980: Nitrate and nitrite reduction. In: Miflin, B. 1.
(ed.), Amino Acids and Derivatives (The Biochemistry of Plants, Vol. 5), pp.
115-168. New York: Academic Press.
Bennetzen, 1. L., 1984: Transposable element Mu 1 is found in multiple copies only
in Robertson's Mutator maize lines. 1. Mol. Appl. Genet. 2, 519-524.
144 J. L. Wray
Biasini, G., Marton, L., 1985: A rapid assay for genetic complementation of nitrate
reductase deficiency via bulk protoplast fusion. Mol. Gen. Genet. 198,353-355.
Birkett, J. A., Rowlands, R. T., 1981: Chlorate resistance and nitrate assimilation in
industrial strains of Penicillium chrysogenum. J. Gen. Microbiol. 123,281-285.
Braaksma, F. J., Feenstra, W. J., 1982a: Isolation and characterization of nitrate
reductase-deficient mutants of Arabidopsis thaliana. Theor. Appl. Genet. 64,
83-90.
Braaksma, F. J., Feenstra, W. J., 1982b: Nitrate reduction in the wild type and a
nitrate reductase deficient mutant of Arabidopsis thaliana. Physiol. Plant. 54,
351-360.
Braaksma, F. J., Feenstra, W. J., 1982 c: Reverse mutants of the nitrate reductase-
deficient mutant B25 of Arabidopsis thaliana. Theor. Appl. Genet. 61, 263-271.
Bray, R. C., 1980: The reaction and the structures of molybdenum centres in
enzymes. In: Meister, A. (ed.), Advances in Enzymology, Vol. 51, pp. 107-166.
New York: Academic Press.
Bright, S. W. J., Norbury, P. B., Franklin, J., Kirk, D. W., Wray, J. L., 1983: A con-
ditional-lethal cnx-type nitrate reductase-deficient barley mutant. Mol. Gen.
Genet. 189,240-244.
Brown, J., Small, I. S., Wray, J. L., 1981: Age-dependent conversion of nitrate
reductase to cytochrome c reductase species in barley leaf extract. Phytochem-
istry 20, 389-398.
Brownlee, A. G., Arst, H. N., Jr., 1983: Nitrate uptake in Aspergillus nidulans and
involvement of the third gene of the nitrate assimilation gene cluster. J. Bac-
teriol. 155, 1138-1146.
Buchanan, R. J., Wray, J. L., 1982: Isolation of molybdenum cofactor-defective
lines of Nicotiana tabacum. Mol. Gen. Genet. 188,228-234.
Buxton, F. R., Radford, A., 1983: Cloning of the structural gene for orotidine
5' -phosphate carboxylase of Neurospora crassa by expression in Escherichia coli.
Mol. Gen. Genet. 190,403-405.
Campbell, A. M., Campbillo-Campbell, A. D., Villaret, D. B., 1985: Molybdate
reduction by Escherichia coli K-12 and its chI mutants. Proc. Nat. Acad. Sci.
U.S.A. 87, 227-231.
Campbell, J. McA., Wray, J. L., 1983: Purification of barley nitrate reductase and
demonstration of nicked subunits. Phytochemistry 22, 2375-2382.
Campbell, J. McA., Small, I. S., Wray, J. L., 1984: Purification of an NADH
cytochrome c reductase from cell-free extracts of barley. Phytochemistry 23,
1391-1395.
Campbell, W. H., 1976: Separation of soybean leaf nitrate reductases by affinity
chromatography. Plant Sci. Lett. 7, 239-247.
Campbell, W. H., 1978: Isolation of NAD (P) H: nitrate reductase from the scu-
tellum of maize. Z. Pflanzenphysiol. 88, 357 -361.
Chen, P. c., Phillips, D. A., 1977: Induction of root nodule senescence by com-
bined nitrogen in Pisum sativum L. Plant Physiol. 59, 440-442.
Clement, C. R., Hopper, M. J., Jones, L. P. H., 1978: The uptake of nitrate by
Lolium perenne from flowing nutrient solution. I. Effect of NO") concentration.
J. Exp. Bot. 29,453-464.
Coddington, A., 1976: Biochemical studies on the nit mutants of Neurospora crassa.
Mol. Gen. Genet. 145, 195-206.
Cove, D. J., 1963: Studies of the biochemistry and genetics of inorganic nitrogen
metabolism in tl1e fungus Aspergillus nidulans. Ph. D. dissertation, University of
Cambridge, England.
Cove, D. J., 1976 a: Chlorate toxicity in Aspergillus nidulans. Studies of mutants

altered in nitrate assimilation. Mol. Gen. Genet. 146, 147-159.
Cove, D. J., 1976b: Chlorate toxicity in Aspergillus nidulans: The selection and
characterisation of chlorate-resistant mutants. Heredity 36, 191-203.
Cove, D. J., 1979: Genetic studies of nitrate assimilation in Aspergillus nidulans.
BioI. Rev. 54, 291-327.
Cove, D. J., Pateman, J. A., 1969: Autoregulation of the synthesis of nitrate
reductase in Aspergillus nidulans. J. Bacteriol. 97, 1374-1378.
Cramer, S. P., Wahl, R, Rajagopalan, K. V., 1981: Molybdenum sites of sulphite
oxidase and xanthine dehydrogenase. A comparison by EXAFS. J. Am. Chern.
Soc. 102, 7721-7727.
Dailey, F. A., Kuo, T., Warner, R L., 1982a: Pyridine nucleotide specificity of
barley nitrate reductase. Plant Physiol. 69, 1196-1199.
Dailey, F. A., Warner, R. L., Somers, D. A., Kleinhofs, A., 1982b: Characteristics of
a nitrate reductase in a barley mutant deficient in NADH nitrate reductase.
Dalling, M. J., Halloran, J. M., Wilson, H. J., 1975: The relation between nitrate
reductase activity and grain nitrogen productivity in wheat. Aust. J. Agric. Res.
26, 1-10.
Deane-Drummond, C. E., 1982: Mechanisms for nitrate uptake into barley
(Hordeum vulgare cv. Fergus) seedlings grown at controlled nitrate concentra-
tions in the nutrient medium. Plant Sci. Lett. 24,79-89.
Deane-Drummond, C. E., 1985: Regulation of nitrate uptake into Chara corallin a
cells via NH t stimulation of NO"3 efflux. Plant Cell Environ. 8, 105 -116.
Deane-Drummond, C. E., 1986: A substrate cycling model for nitrate uptake by
Pisum sativum L. seedlings - a key to sensitivity of response of net flux to sub-
strate and effectors? Plant and Soil, 91, 307-311.
Deane-Drummond, C. E., Glass, A. D. M., 1983: Short term studies of nitrate
uptake into barley plants using ion-specific electrodes and 36CI03. I. Control of
net uptake by N0 3 efflux. Plant Physiol. 73, 100-104.
De Block, M., Herrera-Estrella, L., van Montague, M., Schell, J., Zambryski, P.,
1984: Expression of foreign genes in regenerated plants and in their progeny.
EMBO J. 3, 1681-1689.
Dirks, R., Negrutiu, I., Sidorov, v., Jacobs, M., 1985: Complementation analysis by
somatic hybridisation and genetic crosses of nitrate reductase-deficient mutants
of Nicotiana plumbaginifolia. Evidence for a new category of cnx mutants. Mol.
Gen. Genet. 201, 339-343.
Doddema, H., Telkamp, G. P., 1979: Uptake of nitrate by mutants of Arabidopsis
thaliana disturbed in uptake or reduction of nitrate. Physiol. Plant. 45, 332-338.
Dunn-Coleman, N. S., 1984a: Biochemical characterisation of the molybdenum
cofactor mutants of Neurospora crassa: In vivo and in vitro reconstitution of
NADPH-nitrate reductase activity. Curr. Genet. 8, 581-588.
Dunn-Coleman, N. S., 1984 b: Cloning and preliminary characterization of a
molybdenum cofactor gene of Neurospora crassa. Curro Genet. 8, 589-595.
Dunn-Coleman, N. S., Smarelli, J., Jr., Garrett, R H., 1984: Nitrate assimilation in
eukaryotic cells. Int. Rev. Cytol. 92, 1-50.
Evans, H. J., Nason, A., 1953: Pyridine nucleotide-nitrate reductase from extracts
of higher plants. Plant Physiol. 28, 233-254.
Evans, L. T., 1975: Crop and world food supply, crop evolution, and the origins of
crop physiology. In: Evans, L. T. (ed.), Crop Physiololgy, pp. 1-22. Cam-
bridge: Cambridge University Press.
146 J. L. Wray
Evola, S. V., 1983 a: Chlorate-resistant variants of Nicotiana tabacum L. I. Selection

in vitro and phenotypic characterization of cell lines and regenerated plants.
Mol. Gen. Genet. 189,447-454.
Evola, S. V., 1983 b: Chlorate-resistant variants of Nicotiana tabacum L. II. Para-
sexual genetic characterization. Mol. Gen. Genet. 189,455-457.
Fankhauser, H., Bucher, F., King, P. J., 1984: Isolation of biochemical mutants
using haploid mesophyll protoplasts of Hyoscyamus muticus IV. Biochemical
characterisation of nitrate non-utilising clones. Planta 160, 415-421.
Feenstra, W. J., Jacobsen, E., 1980: Isolation of a nitrate reductase deficient mutant
of Pisum sativum by means of selection for chlorate resistance. Theor. Appl.
Genet. 58, 39-42.
Feenstra, W. J., Jacobsen, E., van Swaay, A. C. P. M., De Visser, A. J. c., 1982:
Effect of nitrate on acetylene reduction in a nitrate reductase deficient mutant
of pea (Pisum sativum L.). Z. Pflanzenphysiol. 105, 471-474.
Ferrari, T. E., Yoder, O. c., Filner, P., 1973: Anaerobic nitrite production by plant
cells and tissues: Evidence for two nitrate pools. Plant Physiol. 51, 423-431.
Fido, R. J., Notton, B. A, 1984: Spinach nitrate reductase: further purification and
removal of 'nicked' sub-units by affinity chromatography. Plant Sci. Lett. 37,
87-91.
Fido, R. J., Hewitt, E. J., Notton, B. A, Jones, O. T. G., Narulhaq-Boyce, A, 1979:
Haem of spinach nitrate reductase: low temperature spectrum and mid-point
potential. FEBS Lett. 99, 180-182.
Foster, S. S. D., Cripps, A. C., Smith-Carington, A, 1982: Nitrate leaching to
groundwater. Philos. Trans. R. Soc. London, Ser. B 296, 477 -489.
Fraley, R. T., Rogers, S. G., Horsch, R. B., Sanders, P. R., Flick, J. S., Adams, S. P.,
Bitter, M. I., Brand, I. A, Fink, C. L., Fry, J. S., Gallupi, G. R., Goldberg, S. B.,
Hoffman, N. I., Woo, S. c., 1983: Expression of bacterial genes in plant cells.
Proc. Nat. Acad. Sci., U.S.A 80, 4803-4807.
Freeling, M., Cheng, D. S.-K., Alleman, M., 1982: Mutant alleles that are altered in
quantitative, organ-specific behaviour. Dev. Genet. 3, 179-196.
Fujihara, S., Yamaguchi, M., 1978: Effects of allopurinol (4' -hydroxypyrazolo
(3,4-d) pyrimidine) on the metabolism of allantoin in soybean plants. Plant
Physiol. 62, l34-138.
Garrett, R. H., Nason, A, 1969: Further purification and properties of Neurospora
nitrate reductase. J. BioI. Chern. 244, 2870-2882.
Gebhardt, c., Schnebli, V., King, P. J., 1981: Isolation of biochemical mutants
using haploid mesophyll protoplasts of Hyoscyamus muticus. II. Auxotrophic
and temperature-sensitive clones. Planta 153, 81- 89.
Gewitz, H.-S., Piefke, J., Vennesland, B., 1981: Purification and characterization of
demolybdonitrate reductase (NADH cytochrome c oxidoreductase) of Chlorella
vulgaris. J. BioI. Chern. 256, 11527-11531.
Gibson, A H., Pagan, J. D., 1977: Nitrate effects on the nodulation of legumes
inoculated with NR-deficient mutants of Rhizobium. Planta 134, 17-22.
Ginns, E. I., Choudary, P. V., Martin, B. M., Winfield, S., Stubblefield, B., Mayor,
J., Merkle-Lehman, D., Murray, G. J., Bowers, L. A., Barranger, J. A, 1984: Iso-
lation of cDNA clones for human ~-glucocerebrosidase using the ).,gtll
expression system. Biochem. Biophys. Res. Commun. 123,574-580.
Giordano, G., Fimmel, A L., Powell, L. M., Haddock, B. A, Barr, G. c., 1981:
Cloning and expression of the chlA gene of Escherichia coli K 12. FEMS Mi-
crobiol. Lett. 12, 61-64.
Glaser, J. H., DeMoss, J. A, 1971: Phenotypic restoration by molybdate of nitrate
reductase activity in chID mutants of Escherichia coli. J. Bacteriol. 108,

854-860.
Glass, A. D. M., Thompson, R. G., Bordelean, L., 1985: Regulation of nitrate influx
in barley. Studies using 13N0"3. Plant Physiol. 77, 379-381.
Glimelius, K., Bonnett, H. T., 1981: Somatic hybridization in Nicotiana: Resto-
ration of photoautotrophy to an albino mutant with defective plastids. Planta
153, 497 -503.
Glimelius, K., Eriksson, T., Grafe, R, Muller, A J., 1978: Somatic hybridization of
nitrate reductase-deficient mutants of Nicotiana tabacum by protoplast fusion.
Physiol. Plant. 44, 273-277.
Glimelius, K., Chen, K., Bonnett, H. T., 1981: Somatic hybridization in Nicotiana:
Segregation of organellar traits among hybrid and cybrid plants. Planta 153,
504-510.
Gottman, K., Schafer, E., 1983: Analysis of phytochrome kinetics in light-grown
Avena sativa L. seedlings. Planta 157, 392-400.
Graf, L. A, Notton, B. A, Hewitt, E. J., 1975: Serological estimation of spinach
nitrate reductase. Phytochemistry 14, 1241-1243.
Grafe, R., Muller, A J., 1983: Complementation analysis of nitrate reductase defi-
cient mutants of Nicotiana tabacum by somatic hybridization. Theor. Appl.
Genet. 66, 127-130.
Gray, 1. C., Jung, S. D., Wildman, S. G., Sheen, S. J., 1974: Origin of Nicotiana
tabacum detected by polypeptide composition of fraction I protein. Nature
(London) 252, 266-267.
Grossman, A R., Barlett, S. G., Schmidt, G. W., Mullet, J. E., Chua, N.-H., 1982:
Optimal conditions for post-translational uptake of proteins by isolated chloro-
plasts. In vitro synthesis and transport of plastocyanin, ferredoxin-NADP+ oxi-
doreductase and fructose-l,6-bisphophatase. J. BioI. Chern. 257, 1558-1563.
Guerrero, M. G., Vega, J. M., Losada, M., 1981: The assimilatory nitrate-reducing
system and its regulation. Annu. Rev. Plant Physiol. 32, 169-204.
Gupta, S. C., Beevers. L., 1984: Synthesis and degradation of nitrite reductase in
pea leaves. Plant Physiol. 75, 251-252.
Gupta, P. P., Gupta, M., Schieder, 0., 1982: Correction of nitrate reductase defect
in auxotrophic plant cells, through protoplast-mediated intergenic gene
transfer. Mol. Gen. Genet. 188, 378-383.
Hamano, T., Oji, Y., Okamoto, S., Mitsuhashi, Y., Matsuki, Y., 1984a: Purification
and characterisation of thiol proteinase as a nitrate reductase-inactivating
factor from leaves of Hordeum distichum L. Plant Cell Physiol. 25, 419-427.
Hamano, T., Oji, Y., Mitsuhashi, Y., Matsuki, Y., Okamoto, S., 1984 b: Action of
thiol proteinase on nitrate reductase in leaves of Hordeum distichum L. Plant
Cell Physiol. 254, 1469-1475.
Hamill, J. D., Pental, D., Cocking, E. C., Muller, A J., 1983: Production of a nitrate
reductase deficient streptomycin resistant mutant of Nicotiana tabacum for
somatic hybridization studies. Heredity 50, 197-200.
Hardy, R. W. F., Holsten, R D., Jackson, E. K., Burns, R C., 1968: The aceylene-
ethylene assay for N2 fixation: Laboratory and field evaluation. Plant Physiol.
43, 1185-1207.
Harper, J. E., 1981 a: Effect of chlorate, nitrogen source and light on chlorate tox-
icity and nitrate reductase activity in soybean leaves. Physiol. Plant 53,
505-510.
Harper, J. E., 1981 b: Evolution of nitrogen oxide(s) during in vivo nitrate reductase
assay of soybean leaves. Plant Physiol. 68, 1488-1493.
148 J. L. Wray
Harper, J. E., Nicholas, D. J. D., 1978: Nitrogen metabolism of soybeans. I. Effect

of tungstate on nitrate utilization, nodulation and growth. Plant Physiol. 62,
662-664.
Hartley, M. J., 1969: Mutation in Aspergillus. Ph. D. Dissertation, University of
Cambridge, England.
Hartley, M. J., 1970: Contrasting complementation patterns in Aspergillus nidulans.
Genet. Res. 16, 123-125.
Heath-Pagliuso, S., Huffaker, R. C., Allard, R. W., 1984: Inheritance of nitrite
reductase and regulation of nitrate reductase, nitrite reductase and glutamine
synthetase isoenzymes. Plant Physiol. 76, 353-358.
Hein, T., Przewozny, T., Schieder, 0., 1983: Culture and selection of somatic
hybrids using an auxotrophic cell line. Theor. Appl. Genet. 64, 119-122.
Helfman, D. M., Feramisco, J. R, Fiddes, J. c., Thomas, G. P., Hughes, S. H., 1983:
Identification of clones that encode chicken tropomyosin by direct immuno-
logical screening of a cDNA expression library. Proc. Nat. Acad. Sci., U.S.A
80,31-35.
Herrera-Estrella, L., De Block, M., Messens, E., Hernalsteens, J. P., van Montague,
M., Schell, J., 1983: Chimaeric genes as dominant selectable markers in plant
cells. EMBO J. 2, 987 -995.
Herrera-Estrella, L., Van den Broeck, R., Maenhaut, R, Van Montagu, M., Schell,
J., Timko, M., Cashmore, A., 1984: Light-inducible and chloroplast-associated
expression of a chimaeric gene introduced into Nicotiana tabacum using a Ti
plasmid vector. Nature (London) 310,115-120.
Hershey, H. P., Colbert, J. T., Lissemore, J. L., Burker, R F., Quail, P. H., 1984:
Molecular cloning of cDNA of Avena phytochrome. Proc. Nat. Acad. Sci.,
U.S.A 81, 2332-2336.
Hewitt, E. J., Notton, B. A, 1980: Nitrate reductase systems in eukaryotic and pro-
karyotic organisms. In: Coughlan, M. (ed.), Molybdenum and Molybdenum-
containing Enzymes, pp. 273-325. Oxford: Pergamon Press.
Hewitt, E. J., Notton, B. A., Rucklidge, G. J., 1977: Formation of nitrate reductase
by recombination of apoprotein fraction from molybdenum deficient plants
with a molybdenum containing complex. J. Less-Common Metals 54, 537 -545.
Highfield, P. E., Ellis, R J., 1978: Synthesis and transport of the small subunit of
chloroplast ribulose bisphosphate carboxylase. Nature (London) 271, 420-424.
Hofstra, J. J., 1977: Chlorate toxicity and nitrate reductase activity in tomato plants.
Physiol. Plant 41, 65-69.
Horner, RD., 1983: Purification and comparison of nit-l and wild-type nitrate
reductases of Neurospora crassa. Biochim. Biophys. Acta 744, 7-15.
Horsch, R. B., Fry, J. E., Hoffmann, N. L., Eichholtz, D., Rogers, S. G., Fraley, R.
T., 1985: A simple and general method for transferring genes into plants.
Science 227, 1229-1231.
Hucklesby, P. D., James, D. M., Banwell, M. J., Hewitt, E. J., 1976: Properties of
nitrite reductase from Cucurbita pepo. Phytochemistry 15, 599-603.
Huffaker, R C., 1982: Biochemistry and physiology of leaf proteins. In: Boulter,
D., Parthier, B. (eds.), Encyclopaedia of Plant Physiology, New Series, Vol.
14A, pp. 370-400. New York - Berlin - Heidelberg: Springer-Verlag.
Hulburt, B. K., Garrett, R. H., 1985: Cloning and regulation of nitrate-inducible
genes in Neurospora crassa. Abst. UCLA Symp. "Molecular Genetics of Fila-
mentous Fungi", p. 1597. New York: Alan R. Liss.
Ida, S., 1977: Purification to homogeneity of spinach nitrite reductase by ferre-
doxin-sepharose affinity chromatography. J. Biochem. 82, 915-918.
Jackson, W. A., Flesher, D., Hageman, R. H., 1973: Nitrate uptake by dark-grown
corn seedlings: some characteristics of apparent induction. Plant Physiol. 51,
120-127.
Jackson, W. A., Kurik, K. D., Yolk, R. J., 1976: Nitrate uptake during recovery from
nitrate deficiency. Plant Physiol. 36, 174-181.
Jacobsen, E., Feenstra, W. J., 1984: A new pea mutant with efficient nodulation in
the presence of nitrate. Plant Sci. Lett. 33, 337 -344.
Jacobsen, E., Braaksma, F. J., Feenstra, W. J., 1984: Determination of xanthine
dehydrogenase activity in nitrate reductase deficient mutants of Pisum sativum
and Arabidopsis thaliana. Z. Pflanzenphysiol. 113, 183 -188.
Jacq, C., Lederer, F., 1974: Cytochrome b 2 from Bakers yeast (L-lactate dehy-
drogenase). A double-headed enzyme. Eur. J. Biochem. 41, 311-320.
Johnson, C. B., Whittington, W. J., Blackwood, G. C., 1976: Nitrate reductase as a
possible predictive test of crop yield. Nature (London) 262, 133 -134.
Johnson, J. L., 1980: The molybdenum cofactor common to nitrate reductase, xan-
thine dehydrogenase and sulfite oxidase. In: Coughlan, M. (ed.), Molybdenum
and Molybdenum-containing Enzymes, pp. 345-383. Oxford: Pergamon Press.
Johnson, J. L., Rajagopalan, K. Y., 1977: Tryptic cleavage of rat liver sulfite
oxidase: Isolation and characterization of molybdenum and heme domains. J.
BioI. Chern. 252,2017-2025.
Johnson, J. L., Rajagopalan, K. Y., 1982: Structural and metabolic relationship
between the molybdenum cofactor and urothione. Proc. Nat. Acad. Sci., U.S.A.
79, 6856-6860.
Johnson, J. L., Hainline, B. E., Rajagopalan, K. Y., 1980: Characterization of the
molybdenum cofactor of sulfite oxidase, xanthine oxidase and nitrate
reductase. Identification of a pteridine as a structural component. J. BioI.
Chern. 255, 1783 -1786.
Johnson, J. L., Hainline, B. E., Rajagopalan, K. Y., Arison, B. H., 1984: Pterin com-
ponent of the molybdenum cofactor. Structural characterization of two fluo-
rescent derivatives. J. BioI. Chern. 259, 5414-5422.
Johnstone, I. J., Hughes, S. G., Clutterbuck, A. J., 1985: Cloning an Aspergillus nid-
ulans developmental gene by transformation. EMBO J. 4, 1307-1311.
Jolly, S. 0., Campbell, W. H., Tolbert, N. E., 1976: NADPH and NADH-nitrate
reductase from soybean leaves. Arch. Biochem. Biophys. 174,431-439.
Jones, P. W., Ni Mhuimhneachain, M., 1985: The activity and stability of wheat
nitrate reductase in vitro. Phytochemistry 24, 385-392.
Kakefuda, G., Duke, S. H., Duke, S. 0., 1983: Differential light induction of nitrate
reductases in greening and photobleached soybean seedlings. Plant Physiol. 73,
56-60.
Ketchum, P. A., Cambier, H. Y., Frazier, W. A. III., Madansky, C. H., Nason, A.,
1970: In vitro assembly of Neurospora mutant assimilatory nitrate reductase
from protein subunits of a Neurospora mutant and the xanthine oxidising and
aldehyde oxidase systems of higher plants. Proc. Nat. Acad. Sci., U.S.A. 66,
1016-1023.
King, J., Khanna, Y., 1980: A nitrate reductase-less variant isolated from
suspension cultures of Datura innoxia (Mill). Plant Physiol. 66, 632-636.
Kinghorn, J. R., Hawkins, A. R., 1982: Cloning and expression in Escherichia coli of
the Aspergillus nidulans arom cluster gene. Mol. Gen. Genet. 186, 145-154.
Kleinhofs, A., Warner, R. L., Muehlbauer, F. J., Nilan, R. A., 1978: Induction and
selection of specific gene mutations in Hordeum and Pisum. Mutat. Res. 51,
29-35.
150 J. L. Wray
Kleinhofs, A, Kuo, T., Warner, R L., 1980: Characterization of nitrate reductase-

deficient barley mutants. Mol. Gen. Genet. 177,421-425.
Kleinhofs, A, Taylor, J., Kuo, T. M., Somers, D. A, Warner, R. L., 1983: Nitrate
reductase genes as selectable markers for plant cell transformation. In:
Lurquin, P. F., Kleinhofs, A (eds.), Genetic Engineering in Eukaryotes, pp.
215-231. New York: Plenum.
Kleinhofs, A, Warner, R. L., Narayanan, K. R, 1985: Current progress towards an
understanding of the genetics and molecular biology of nitrate reductase in
higher plants. In: Miflin, B. J. (ed.), Oxford Surveys of Plant Molecular and
Cell Biology, Vol. 2, pp. 91-121. Oxford: Oxford University Press.
Kramer, S., Hageman, R. V., Rajagopa1an, K. V., 1984: In vitro reconstitution of
nitrate reductase activity of the Neurospora crassa mutant nit-I: specific incor-
poration of molybdopterin. Arch. Biochem. Biophys. 233, 821-829.
Kuh1meier, C. J., Logtenberg, T., Stoorvoge1, W., van Heughten, H. A A, Borrias,
W. E., van Arkel, G. A, 1984: Cloning of nitrate reductase genes from the cya-
nobacterium Anacystis nid_ulans. .1. Bacteriol. 159, 36-41.
Kuo, T., Kleinhofs, A, Warner, R L., 1980: Purification and partial characteri-
sation of nitrate reductase from barley leaves. Plant Sci. Lett. 17,371-381.
Kuo, T., Kleinhofs, A, Somers, D., 'Warner, R. L., 1981: Antigenicity of nitrate
reductase deficient mutants in Hordeum vulgare L. Mol. Gen. Genet. 181, 20-23.
Kuo, T. M., Somers, D. A, Kleinhofs, A, Warner, R. L., 1982: NADH-nitrate
reductase in barley leaves. Identification and amino-acid composition of
subunit protein. Biochem. Acta 708, 75-81.
Kuo, T. M., Kleinhofs, A., Somers, D. A., Warner, R. L., 1984: Nitrate reductase-
deficient mutants in barley: enzyme stability and peptide mapping. Phy,tochem-
istry 23, 229-232. -
Laan, P. H., Van der, Oostindier-Braaksma, F. J., Feenstra, W. J., 1971: Chlorate
resistant mutants of Arabidopsis thaliana. Arabidopsis Inf. Servo 8, 22.
Lahav, E., Harper, J. E., Hageman, R H., 1976: Improved soybean growth in urea
with pH buffered by a carboxy resin. Crop Sci. 16,325-328.
Lancaster, J. R, Vega, J. M., Kamin, H., Orme-Johnson, N. R., Orme-Johnson, W.
H., Kreuger, R. J., Siegel, L. M., 1979: Identification of the iron-sulphur centre
of spinach ferredoxin-nitrite reductase as a tetranuc1ear centre and preliminary
EPR studies of mechanism. J. BioI. Chern. 254, 1268-1272.
Larkin, P. J., Scowcroft, W. R, 1981: Somac1onal variation - a novel source of
variability from cell cultures for plant improvement. Theor. Appl. Genet. 60,
197-214.
Lazar, G. B., Fankhauser, H., Potrykus, I., 1983: Complementation analysis of a
nitrate reductase deficient Hyoscyamus muticus cell line by somatic hybridi-
zation. Mol. Gen. Genet. 189,359-364.
Le, K. H. D., Lederer, F., 1983: On the presence of a heme-binding domain homol-
ogous to cytochrome b s in Neurospora crassa assimilatory nitrate reductase.
EMBO J. 2, 1909-1914.
Lee, K. Y., Pan, S.-S., Erickson, P., Nason, A, 1974: Involvement of molybdenum
and iron in the in vitro assembly of assimilatory nitrate reductase utilising
mutant nitl. J. BioI. Chern. 249, 3941-3952.
Lijlestrom, S., Aberg, B., 1966: Studies on the mechanism of chlorate toxicity.
Kungl. Lantbrukshogkolans Ann. 32, 93 -1 07.
MacDonald, D. W., Cove, D. J., 1974: Studies on temperature-sensitive mutants
affecting the assimilatory nitrate reductase of Aspergillus nidulans. Eur. J.
Biochem.47,107-110.
MacDonald, D. W., Cove, D. J., Coddington, A., 1974: Cytochrome c reductase

from wild type and mutant strains of Aspergillus nidulans. Mol. Gen. Genet.
128, 187-199.
Magee, P. N., 1982: Nitrogen as a potential health hazard. Philos. Trans. R. Soc.
London, Ser. B 296, 543-550.
Maliga, P., 1976: Isolation of mutants from cultured cells. In: Dudits, D., Farkas,
G. L., Maliga, P. (eds.), Cell Genetics in Higher Plants, pp. 59-76. Budapest:
Akademiai Kiado.
Manhart, J. R, Wong, P. P., 1980: Nitrate effects on nitrogen fixation (acetylene
reduction). Plant Physiol. 65, 502-505.
Maretski, A., De La Cruz, A., 1967: Nitrate reductase in sugarcane tissues. Plant
Cell Physiol. 7, 605-611.
Marton, L., Dung, T. M., Mendel, R. R, Maliga, P., 1982a: Nitrate reductase defi-
cient cell lines from haploid protoplast cultures of Nicotiana plumbaginifolia.
Mol. Gen. Genet. 182,301-304.
Marton, L., Sidorov, V., Biasini, G., Maliga, P., 1982 b: Complementation in
somatic hybrids indicates four types of nitrate reductase deficient lines in Nico-
tiana plumbaginifolia. Mol. Gen. Genet. 187, 1-3.
Marton, L., Biasini, G., Maliga, P., 1985: Co-segregation of nitrate reductase
activity and normal regeneration ability in selfed sibs of Nicotiana plumbagini-
folia somatic hybrids, heterozygotes for nitrate reductase deficiency. Theor.
Appl. Genet. 70, 340-344
McClure, P. R, Israel, D. W., 1979: Transport of nitrogen in the xylem of soybean
plants. Plant Physiol. 64, 411-416.
Mendel, R-R, 1983: Release of molybdenum co-factor from nitrate reductase and
xanthine oxidase by heat treatment. Phytochemistry 22,817-819.
Mendel, R R, Muller, A. J., 1976: A common genetic determinant of xanthine
dehydrogenase and nitrate reductase in Nicotiana tabacum. Biochem. Physiol.
Pflanzen. 170, 538-541.
Mendel, R R., Muller, A. J., 1979: Nitrate reductase-deficient mutant cell lines of
Nicotiana tabacum. Further biochemical characterization. Mol. Gen. Genet.
177, 145-153.
Mendel, R.-R, Muller, A. J., 1980: Comparative characterisation of nitrate
reductase from wild-type and molybdenum cofactor-defective cell cultures of
Nicotiana tabacum. Plant Sci. Lett. 18, 277 -288.
Mendel, R-R., Muller, A. J., 1985: Repair in vitro of nitrate reductase-deficient
tobacco mutants (cnxA) by molybdate and by molybdenum cofactor. Planta
163, 370-375.
Mendel, R-R, Alikulov, Z. A., Lvov, N. P., Muller, A. J., 1981: Presence of the
molybdenum-cofactor in nitrate reductase-deficient mutant cell lines of Nico-
tiana tabacum. Mol. Gen. Genet. 181,395-399.
Mendel, R R., Alikulov, Z. A., Muller, A. J., 1982: Molybdenum cofactor in nitrate
reductase-deficient tobacco mutants. III. Induction of cofactor synthesis by
nitrate. Phytochemistry 27, 95-101.
Mendel, R-R, Buchanan, R J., Wray, J. L., 1984: Characterization of a new type
of molybdenum cofactor-mutant in cell cultures of Nicotiana tabacum. Mol.
Gen. Genet. 195, 186-189.
Mendel, R-R., Kirk, D. W., Wray, J. L., 1985: Assay of molybdenum cofactor of
barley. Phytochemistry 24, 1631-1634.
Mendel, R R., Marton, L., Muller, A. J., 1986: Comparative biochemical character-
152 J. L. Wray
ization of mutants at the nitrate reductase/molybdenum cofactor loci cnxA,

cnxB and cnxC of two Nicotiana species. Plant Sci. 43, 125-129.
Meriwether, L. S., Marzluff, W. F., Hodgson, W. G., 1966: Molybdenum-thiol com-
plexes as models for molybdenum bound in enzymes. Nature (London) 212,
465-467.
Miller, B. L., Huffaker, R c., 1981: Partial purification and characterization of
endoproteinases from senescing barley leaves. Plant Physiol. 68, 930-936.
Minotti, P. L., Williams, D. C., Jackson, W. A., 1968: Nitrate uptake and reduction
as affected by calcium and potassium. Soil Sci. Soc. Am. 32, 692-698.
Morelli, G., Nagy, F., Fraley, R T., Rogers, S. C., Chua, N.-H., 1985: A short con-
served sequence is involved in the light-inducibility of a gene encoding ribulose
1,5-bisphosphate carboxylase small subunit of pea. Nature (London) 315,
200-204.
Muller, A. J., 1983: Genetic analysis of nitrate reductase-deficient tobacco plants
regenerated from mutant cells. Evidence for duplicate structural genes. Mol.
Gen. Genet. 192, 275-281.
Muller, A. J., Grafe, R., 1978: Isolation and characterization of cell lines of Nico-
tiana tabacum lacking nitrate reductase. Mol. Gen. Genet. 161, 67 - 76.
Mulvaney, C. S., Hageman, R H., 1984: Acetaldehyde oxime, a product formed
during in vivo nitrate reductase assay of soybean leaves. Plant Physiol. 76,
118-124.
Munns, D. N., 1968: Nodulation of Medicago sativa in solution culture. III. Effects
of nitrate on root hairs and infection. Plant and Soil 29, 33-47.
Murphy, J. M., Imbrie, C. W., 1981: Induction and characterization of chlorate-
resistant strains of Rosa damascena cultured cells. Plant Physiol. 67,910-916.
Nakagawa, H., Yonemura, Y., Yamamoto, H., Sato, T., Ogura, N., Sato, R, 1985:
Spinach nitrate reductase. Purification, molecular weight and subunit compo-
sition. Plant Physiol. 77, 124-128.
Narayanan, K. R, Somers, D. A., Kleinhofs, A., Warner, R L., 1983: Nature of
Cytochrome c reductase deficient mutants in barley. Mol. Gen. Genet. 190,
222-226.
Narayanan, K. R, Muller, A. J., Kleinhofs, A., Warner, R L., 1984: In vitro recon-
stitution of NADH: nitrate reductase in nitrate reductase deficient mutants of
barley. Mol. Gen. Genet. 197, 358-362.
Naslin, L., Spyridakis, A., Labeyrie, F., 1973: A study of several bonds hypersen-
sitive to proteases in a complex flavohaemoenzyme, yeast cytochrome b 2 Modi-
fication of their reactivity with ligand-induced conformational transitions. Eur.
J. Biochem. 34, 268-283.
Nason, A., Lee, K.-Y., Pan, S.-S., Ketchum, P. A., Lamberti, A., Davies, J., 1971: In
vitro formation of assimilatory reduced nicotinamide adenine dinucleotide
phosphate: nitrate reductase from a Neurospora mutant and a component of
molybdenum-enzymes. Proc. Nat. Acad. Sci., U.S.A. 68, 3242-3246.
Negrutiu, I., Dirks, R., Jacobs, M. L., 1983: Regeneration of fully nitrate reductase-
deficient mutants from protoplast culture of Nicotiana plumbaginifolia (Viviani).
Theor. Appl. Genet. 66, 341-347.
Nelson, R. S., Ryan, S. A., Harper, J. E., 1983: Soybean mutants lacking consti-
tutive nitrate reductase activity. I. Selection and initial plant characterization.
Nelson, R S., Streit, L., Harper, J. E., 1984: Biochemical characterization of nitrate
and nitrite reduction in the wild type and a nitrate reductase mutant of soybean.
Physiol. Plant 61, 384-390.
Notton, B. A., Hewitt, E. J., 1971 a: Incorporation of radioactive molybdenum into

protein during nitrate reductase formation and effect of molybdenum on nitrate
reductase and diaphorase activities of spinach (Spinacea oleracea L). Plant Cell
Physiol. 12,465-477.
Notton, B. A., Hewitt, E. J., 1971 b: The role of tungsten in the inhibition of nitrate
reductase activity in spinach (Spinacea oleracea) leaves. Biochem. Biophys. Res.
Commun. 44, 702-710.
Notton, B. A., Hewitt, E. J., 1979: Structure and properties of higher plant nitrate
reductase especially Spinacea oleracea L. In: Hewitt, E. J., Cutting, C. V. (eds.),
Nitrogen Assimilation in Plants, pp. 227-244. New York: Academic Press.
Notton, B. A., Fido, R. J., Hewitt, E. J., 1977: The presence of a functional haem in
a higher plant nitrate reductase. Plant Sci. Lett. 8, 165 -170.
Notton, B. A., Fido, R. J., Galfre, G., 1985: Monoclonal antibodies to a higher plant
nitrate reductase: differential inhibition of enzyme activity. Planta 165, 114-119.
Oh, J. Y., Warner, R. L., Kleinhofs, A., 1980: Effect of nitrate reductase deficiency
upon growth yield and protein in barley. Crop Sci. 20, 487 -490.
Oostindier-Braaksma, F. J., 1970: A chlorate resistant mutant of Arabidopsis
thaliana. Arabidopsis Inf. Servo 7, 24.
Oostindier-Braaksma, F. J., Feenstra, W. J., 1972: Chlorate resistant mutants of
Arabidopsis thaliana II. Arabidopsis Inf. Servo 7, 9-10.
Oostindier-Braaksma, F. J., Feenstra, W. J., 1973: Isolation and characterization of
chlorate-resistant mutants of Arabidopsis thaliana. Mutat. Res. 19, 175-185.
O'Reilly, C., Shepherd, N. S., Pereira, A., Schwarz-Sommer, Z., Bertram, I.,
Robertson, D. S., Peterson, P. A., Saedler, H., 1985: Molecular cloning of the
a 1 locus of Zea mays using the transposable elements En and Mu 1. EMBO J. 4,
877-882.
Orihue1-Iranzo, B., Campbell, W. H., 1980: Development of NAD (P) H: and
NADH: nitrate reductase activities in soybean cotyledons. Plant Physiol. 65,
595-599.
Pan, S-S., Nason, A., 1978: Purification and characterization of homogeneous as-
similatory reduced nicotinamide adenine dinucleotide phosphate-nitrate
reductase from Neurospora crassa. Biochim. Biophys. Acta 523, 297-313.
Paszowski, J., Shillito, R. D., Saul, M., Mandak, V., Hohn, T., Hohn, P., Potrykus,
I., 1984: Direct gene transfer to plants. EMBO J. 3, 2712-2722.
Pateman, J. A., Cove, D. J., Rever, B. M., Roberts, D. B., 1964: A common co-factor
for nitrate reductase and xanthine dehydrogenase which also regulates the syn-
thesis of nitrate reductase. Nature (London) 201, 58-60.
Pateman, J. A., Rever, B. M., Cove, D. J., 1967: Genetic and biochemical studies of
nitrate reduction in Aspergillus nidulans. Biochem. J. 104, 103 -111.
Pental, D., Cooper-Bland, S., Harding, K., Cocking, E. C., Muller. A. J., 1982: Cul-
tural studies on nitrate reductase deficient Nicotiana tabacum mutant proto-
p1asts. Z. Pflanzenphysiol. 105,219-227.
Pienkos, P. T., Shah, V. K., Brill, W. J., 1977: Molybdenum co-factors from
molybdo-enzymes and in vitro reconstitution of nitrogenase and nitrate
reductase. Proc. Nat. Acad. Sci., U.S.A. 74, 5468-5471.
Qureshi, J. A., Buchanan, R. J., Wray, J. L., 1982: Isolation and partial character-
ization of a chlorate-resistant cell line of Nicotiana tabacum possessing nitrate
reductase activity. Proc. Internat. Symp. "Nitrate Assimilation - Molecular
and Genetic Aspects", Gatersleben, DDR, P. 22.
Rao, K. P., Rains, D. W., 1976: Nitrate absorption by barley. I. Kinetics and ener-
getics. Plant Physiol. 57, 55-58.
154 J. L. Wray
Redinbaugh, M. G., Campbell, W. H., 1981: Purification and characterization of

NAD(P)H: nitrate reductase and NADH: nitrate reductase from corn roots. Plant
Physiol. 68, 115 -120.
Redinbaugh, M. G., Campbell, W. H., 1983: Purification of squash NADH: nitrate
reductase by zinc chelate affinity chromatography. Plant Physiol. 71, 205-207.
Redinbaugh, M. G., Campbell, W. H., 1985: Quaternary structure and composition of
squash NADH: nitrate reductase. J. BioI. Chem. 260, 3380-3385.
Reed, A. J., Hageman, R. H., 1980a: Relationship between nitrate uptake, flux, and
reduction and the accumulation of reduced nitrogen in maize (Zea mays L). I.
Genotypic variation. Plant Physiol. 66, 1179-1183.
Reed, A. J., Hageman, R. H., 1980b: Relationship between nitrate uptake, flux, and
reduction and the accumulation of reduced nitrogen in maize (Zea mays L.). Plant
Physiol. 66, 1184-1189.
Relimpio, A. Ma., Aparicio. P. J., Paneque, A., Losada, M., 1971: Specific protection
against inhibitors of the NADH-nitrate reductase complex from spinach. FEBS
Lett. 17,226-230.
Rever, B. M., 1966: Biochemical and genetical studies of inorganic nitrogen metab-
olism in Aspergillus nidulans. Ph. D. Dissertation, University of Cambridge,
England.
Reynolds, P. H. S., Boland, M. J., Blevins, D. G., Schubert, K. R., Randall, D. D.,
1982: Enzymes of ureide biogenesis in developing soybean nodules. Plant
Physiol. 69, 1334-1338.
Robertson, D. S., 1978: Characterization of a mutator system in maize. Mutant Res.
51,21-28.
Robin, P., Streit, L., Campbell, W. H., Harper, J. E., 1985: Immunochemical charac-
terization of nitrate reductase forms from wild-type (cv. Williams) and nrl mutant
soybean. Plant Physiol. 77, 232-236.
Ruiz-Herrera, J., Showe, M. K., DeMoss, J. A., 1969: Nitrate reductase complex of
Escherichia coliK-12: Isolation and characterization of mutants unable to reduce
nitrate. J. Bacteriol. 97,1291-1297.
Ryan, S. A., Nelson, R. S., Harper, J. E., 1983: Soybean mutants lacking constitutive
nitrate reductase activity. I. Nitrogen assimilation, chlorate resistance and inher-
itance. Plant Physiol. 72, 510-514.
Savage, A. D., King, J., Gamborg, O. L., 1979: Recovery of a pantothenate auxotroph
from a cell suspension culture of Datura innoxia. Mill. Plant. Sci. Lett. 16,
367-376.
Schechtman, M. G., Yanofsky, C., 1983: Structure of the trifunctional trp-l gene from
Neurospora crassa and its aberrant expression in Escherichia coli. J. Mol. Appl.
Genet. 2, 83-99.
Schell, J., van Montagu, M. 1983: The Ti plasmids as natural and as practical gene
vectors for higher plants. Biotechnology 1,175-180.
Schieder, 0., Vasil, I. K., 1980: Protoplast fusion and somatic hybridisation. In: Vasil,
I. K. (ed.), Recent Advances in Plant Cell and Tissue Culture. Int. Rev. Cytol.
Suppl.118, pp. 21-46. New York: Academic Press.
Schrader, L. E., Ritenour, G. L., Eilrich, G. L., Hageman, R. H., 1968: Some character-
istics of nitrate reductase from higher plants. Plant Physiol. 43, 930-940.
Serra, J. L., Ibarlucea, J. M., Arizmerdi, J. M., Llama, M. J., 1982: Purification and
properties of the assimilatory nitrite reductase from barley Hordeum vulgare
leaves. Biochem. J. 201,167-170.
Shah, D. M., Hightower, R. c., Meagher, R. B., 1983: Genes encoding actin in higher
plants: Intron positions are highly conserved but the coding sequences are not. J.
Mol. Appl. Genet. 2, 111-126.
Shah, V. K., Brill, W. J., 1977: Isolation of an iron-molybdenum cofactor from
nitrogenase. Proc. Nat. Acad. Sci., U.S.A. 74, 3249-3252.
Shaner, D. L., Boyer, J. S., 1976: Nitrate reductase activity in maize (Zea mays L.)
leaves. I. Regulation by nitrate flux. Plant Physiol. 58, 499-504.
Shen, T. C., Funkhauser, E. A., Guerrero, M. G., 1976: NADH- and NAD(P)H-
nitrate reductases in rice seedlings. Plant Physiol. 58, 292-297.
Shepard, J. F., Bidney, D., Barsby, T., Kemble, R., 1983: Genetic transfer in plants
through interspecific protoplast fusion. Science 219, 683-688.
Shumny, V. K., Tokarev, B. I., 1982: Genetic control of nitrate reductase activity in
barley. Proc. IVth IntI. Barley Symp.
Small, I. S., Gray, J. C., 1984: Synthesis of wheat leaf nitrite reductase de novo fol-
lowing induction with nitrite. Eur. J. Biochem. 145, 291-297.
Small, I. S., Wray, J. L., 1980: NADH nitrate reductase and related NADH
cytochrome c reductase species in barley. Phytochemistry 19, 387 -394.
Smarrelli, J., Campbell, W. H., 1981: Immunological approach to structural com-
parisons of assimilatory nitrate reductases. Plant Physiol. 68, 1226-1230.
Solomons on, L. P., Vennesland, B., 1972: Nitrate reductase and chlorate toxicity in
Chlorella vulgaris Beyerinck. Plant Physiol. 50, 421-424.
Solomonson, L. P., Barber, M. J., Howard, W. D., Johnson, J. L., Rajagopalan, K.
V., 1984: Electron paramagnetic resonance studies on the molybdenum center
of assimilatory NADH: nitrate reductase from Chlorella vulgaris. J. BioI. Chern.
259, 849-853.
Somers, D. A., Kuo, T., Kleinhofs, A., Warner, R. L., 1982: Barley nitrate reductase
contains a functional cytochrome b 557 Plant Sci. Lett. 24, 261-265.
Somers, D. A., Kuo, T. M., Kleinhofs, A., Warner, R. L., 1983: Nitrate reductase
deficient mutants in barley. Plant Physiol. 71, 145-149.
Sood, A. K., Pereira, D., Weissman, S. M., 1981: Isolation and partial nucleotide
sequence of a cDNA clone for human histocompatibility antigen H2A-B by use
of an oligodeoxynucleotide primer. Proc. Nat. Acad. Sci., U.S.A. 78, 616-620.
Sorger, G. J., Giles, N. H., 1965: Genetic control of nitrate reductase in Neurospora
crassa. Genetics 52, 777 - 788.
Southerland, W. H., Rajagopalan, K. V., 1978: Domain interactions in oxidised and
reduced forms of rat liver sulfite oxidase. J. BioI. Chern. 253, 8753-8758.
Srivastava, H. S., 1980: Regulation of nitrate reductase activity in higher plants.
Phytochemistry 19, 725-733.
Steffen, A., Schieder, 0., 1984: Biochemical and genetical characterization of
nitrate reductase deficient mutants of Petunia. Plant Cell Rep. 3, 134-137.
Stewart, G. R., Orebamjo, T. 0., 1979: Some unusual characteristics of nitrate
reductase in Erythrina senegalensis DC. New Phytol. 83, 311-319.
Stewart, V., MacGregor, C. H., 1982: Nitrate reductase in Escherichia coli K-12:
Involvement of chIC, chlE and chlG loci. J. Bacteriol. 151, 788-799.
Strauss, A., Bucher, F., King, P. J., 1981: Isolation of biochemical mutants using
haploid mesophyll protoplasts of Hyoscyamus muticus. I. A NO"3 non-utilising
clone. Planta, 153, 75-80.
Streeter, J. G., 1985: Nitrate inhibition of legume nodule growth and activity. I. Long
term studies with a continuous supply of nitrate. Plant Physiol. 77, 321-324.
Streit, L., Nelson, R. S., Harper, J. E., 1985: Nitrate reductases from wild-type and
llfl-mutant soybean (Glycine max (L.) Merr.) leaves. Plant Physiol. 78, 80-84.
Strommer, J. N., Hake, S., Bennetzen, J. L., Taylor, W. C., Freeling, M., 1982: Regu-
156 J. L. Wray
1atory mutants of the maize Adh 1 gene caused by DNA mutations. Nature
(London) 300, 542-544.
Suggs, S. Y., Wallace, R B., Hirose, T., Kawashima, E. H., Itakura, K., 1981: Use of
synthetic oligonucleotides as hybridization probes: Isolation of cloned cDNA
sequences for ~2-microg10bulin. Proc. Nat. Acad. Sci., U.S.A 78, 6613-6617.
Taylor, J. L., Bedbrook, J. R, Grant, F. J., Kleinhofs, A., 1983: Reconstitution of
plant nitrate reductase by Escherichia coli extracts and the molecular cloning of
the chlA gene of Escherichia coli K12. J. Mol. Appl. Genet. 2, 261-271.
Thomas, R. J., Schrader, L. E., 1981: Ureide metabolism in higher plants. Phyto-
chemistry 20, 361-371.
Thompson, S. T., Kass, K. H., Stellwagen, E., 1975: Blue dextran-Sepharose: An
affinity column for the dinucleotide fold in proteins. Proc. Nat. Acad. Sci.,
U.S.A 72, 669-672.
Tilburn, J., Scazzochio, C., Taylor, G. T., Zabicky-Zissman, J. H., Lockington, R
A, Davies, W. R, 1984: Transformation by integration in Aspergillus nidulans.
Gene 26, 205-221.
Tokarev, B. I., Shumny, U. K., 1977: Clarification of barley mutants with lowered
nitrate reductase activity after treatment of the grain with ethy1methanesul-
phonate. Genetika 13,2097-2103.
Tomsett, A B., Cove, D. J., 1979: Deletion mapping of the niiA niaD gene region of
Aspergillus nidulans. Genet. Res. 34, 19-32.
Tomsett, A. B., Garrett, R H., 1980: The isolation and characterization of mutants
defective in nitrate assimilation in Neurospora crassa. Genetics 95, 649-660.
Tomsett, A. B., Dunn-Coleman, N. S., Garrett, R. H., 1981: The regulation of
nitrate assimilation in Neurospora crassa: The isolation and genetic analysis of
nmr-1 mutants. Mol. Gen. Genet. 182,229-233.
Triplett, E. W., 1985: Intercellular nodule localization and nodule-specificity of
xanthine dehydrogenase in soybean. Plant Physiol. 77, 1004-1009.
Vapnek, D., Hautala, J. A, Jacobson, J. W., Giles, N. H., Kushner, S. R., 1977:
Expression in Escherichia coli K-12 of the structural gene for catabolic dehy-
drogenase of Neurospora crassa. Proc. Nat. Acad. Sci. USA 74,3508-3512.
Vega, J. M., 1972: NADH-nitrate reductase de Ch10rella. Nouvelle contribution a
l'etude de ses properties. Physiol. Veg. 10,637-652.
Vega, J. M., Kamin, H., 1977: Spinach nitrite reductase. Purification and properties
of a siroheme-containing iron-sulphur centre. J. BioI. Chern. 252, 896-909.
Vierstra, R D., Quail, P. H., 1983: Purification and initial characterization of 124K
dalton phytochrome from Avena. Biochem. 22, 2498-2505.
Wahl, R C., Hageman, R V., Rajagopalan, K. V., 1984: The relationship of Mo,
molybdopterin, and the cyanolyzable sulphur in the Mo cofactor. Arch.
Biochem. Biophys. 230, 264-273.
Wallace, W., Oaks. A, 1984: Role of proteinases in the regulation of nitrate
reductase. In: Dalling, M. J. (ed.), Plant Proteases. Boca Raton: CRC Press.
Wallace, R. B., Johnson, N. J., Hirose, T., Mijake, M., Kawashima, E. H., Itakura,
K., 1981: The use of synthetic nucleotides as hybridization probes. II. Hybridi-
zation of oligonucleotides of mixed sequence to rabbit ~-globin DNA Nuc.
Acids Res. 9, 879-894.
Wallin, A, Glimelius, K., Eriksson, T., 1979: Formation of hybrid cells by transfer
of nuclei via. fusion of miniprotoplasts from cell lines of nitrate reductase defi-
cient tobacco. Z. Pflanzenphysiol. 91, 89-94.
Warner, R L., Kleinhofs, A, 1981: Nitrate utilization by nitrate reductase-deficient
barley mutants. Plant Physiol. 67, 740-743.
Warner, R. L., Lin, C. J., Kleinhofs, A., 1977: Nitrate reductase-deficient mutants in
barley. Nature (London) 269, 406-407.
Warner, R. L., Kleinhofs, A., Muehlbauer, F. J., 1982: Characterization of nitrate
reductase-deficient mutants in pea. Crop Sci. 22, 389-393.
Weaver, R. J., 1942: Some responses of the bean plant to chlorate and perchlorate
ions. Plant Physiol. 17, 123-128.
Wilkerson, J. 0., Janick, P. A., Siegel, L. M., 1983: Siroheme-Fe4S4 interactions in
spinach nitrite reductase (NiR). Fed. Proc. Fed. Am. Soc. Exp. BioI. 42, 2060.
Wilkinson, W. B., Greene, L. A., 1982: The water industry and the nitrogen cycle.
Philos. Trans. R. Soc. (London), Ser. B 296,459-475.
Wray, J. L., Filner, P., 1970: Structural and functional relationships of enzyme
activities induced by nitrate in barley. Biochem. J. 119,715-725.
Wray, J. L., Kirk, D. W., 1981: Inhibition of NADH-nitrate reductase degradation
in barley leaf extracts by leupeptin. Plant Sci. Lett. 23, 207 -213.
Wray, J. L., Steven, B., Kirk, D. W., Bright, S. W. J. 1985: A conditional-lethal
molybdopterin-defective mutant of barley. Mol. Gen. Genet. 201, 462-466.
Xuan, L. T., Grafe, R., Muller, A. J., 1983: Complementation of nitrate reductase
deficient mutants in somatic hybrids between Nicotiana species. In: Potrykus,
I., Harms, C. T., Hinnen, A., Hutter, R., King, P. J., Shillito, R. D. (eds.), Proto-
plasts 1983 (Poster Proceedings), pp. 76-77. Basel: Birkhauser.
Yatsunami, K., Pandya, B. V., Oprian, D. D., Khorana, H. G., 1985: cDNA-derived
amino acid sequence of the y-subunit of GTPase from bovine rod outer seg-
ments. Proc. Nat. Acad. Sci., U.S.A. 82, 1936-1940.
Young, R. A., Davies. R. W., 1983 a: Efficient isolation of genes by using antibody
probes. Proc. Nat. Acad. Sci., U.S.A. 80, 1194-1198.
Young, R. A., Davies, R. W., 1983 b: Yeast RNA polymerase II genes: Isolation
with antibody probes. Science 222, 778- 782.
Zabala, G., Filner, P., 1980: Reduction of chlorate by nitrate reductase of tobacco
XD cells. 15th Ann. Report MSU/DOE Plant Res. Lab., pp. 72-73.
Zambryski, P., Joos, H., Genetello, c., Leemans, J., van Montagu, M., Schell, J.,
1983: Ti plasmid vector for the introduction of DNA into plant cells without
alteration of their normal regeneration capacity. EMBO J. 2,2143-2150.
Chapter 6
Plant Genetic Approaches to Symbiotic Nodulation and

Nitrogen Fixation in Legumes
Peter M. Gresshoff and Angela C. Delves
Botany Department, Australian National University, Canberra ACT 2600, Australia
With 6 Figures
Contents
I.Introduction
II.A General Description of Legume Nodule Ontogeny
III. The Parasponia-Bradyrhizobium Symbiosis
IV. Biochemical and Molecular Analysis of Plant Functions
V. Gene-for-Gene Aspects of Nodulation
VI. Existing Plant Variation in Symbiotic Nitrogen Fixation
VII. Existing Single Locus Variation for Nodulation-Nitrogen Fixation
VIII. Induced Mutation in Symbiotic Characters
A. Pea and Chickpea Mutants
B. Soybean Nodulation Mutants
IX. Conclusions
X. References
I. Introduction
We happily accepted the invitation to write this chapter because we have

experienced the development of new approaches to research into the ge-
netics, microbiology and biochemistry of symbiotic nitrogen fixation. We
were confronted with a broad range of advances in the biochemical ge-
netics of Rhizobium, especially relating to the genetic elements controlling
the key symbiotic phenotypes: nitrogen fixation (the nif and fix genes) and
nodulation (controlled by the nod genes). Likewise there has been a major
application of plant molecular biology to the analysis of nodule-specific or
nodule-enhanced plant gene products (Lee et al., 1983; Fuller et ai., 1983;
Govers et al., 1985; and Kaninakis and Verma, 1985). This major class of
proteins or nodulins allows the molecular visualisation of the plant
genome's contribution to the symbiosis (Verma et al., 1985).
160 P. M. Gresshoff and A. C. Delves
The first volume of Plant Gene Research deals with the recent advances
in plant-bacterial interactions (see also references by Dazzo and Gardiol,
1984; Miflin and Cullimore, 1984; Rolfe and Shine, 1984; Verma and
Nadler, 1984). However, we feel that the molecular analyses of the bac-
terial genome (see Djordjevic et aI., 1983; Fischer and Hennecke, 1984;
Schmidt et al., 1984; Schofield et al., 1983, 1984; Masterson et al., 1985)
and the nodulin sequence analysis have largely outstripped the level of
understanding at the phenotypic level of whole plant biology. Even our
knowledge of cellular processes is limited, although it is possible now to
use a well-defined bacterial mutant to probe the plant genotype. For
example, one can use lacZ fusions to bacterial nodulation promoters to
evaluate the excretion of plant signals during early nodulation (Olson et
al., 1985).
Numerous plant functions are involved in nodule initiation and nodule
function. Some of these are under direct bacterial control, others are devel-
opmentally regulated by endogenous plant factors. Nearly every plant
function is integrated in the symbiotic process. However, we know little
about essential mechanisms such as growth regulator activity, translo-
cation, cell wall biosynthesis, developmental gradients, plant genome
structure, gas exchange and especially gene regulation. No wonder that we
really know so little about symbiotic nitrogen fixation. Perhaps by
focussing on two easily recognisable phenotypes and assayable characters
of development, like nodule development and nitrogen fixation, one can
study phenomena which relate to plants in general, rather than the nodu-
lation response per se.
Nodulation in legumes by Rhizobium offers a unique experimental
system to study plant development as it is possible to isolate plant mutants
altered in the symbiotic process, without their fitness being affected
(Carroll et al., 1985 a, b; Gresshoff et al., 1985 b). This is due to the ability
of legumes to use alternative modes of nitrogen acquisition such as nitrate
reduction. This flexibility of course aids the investigation of both plant and
bacterial mutants. In contrast, studies of flower initiation, root devel-
opment, phytohormone biochemistry and/or fruit ripening are experimen-
tally more complex, particularly if one requires a genetic approach which
is to lead into analytial biochemistry of the organism and its development.
Thimann in 1936 stated that it was 50 years since the original dis-
coveries of the nitrogen fixing root nodule, and the first successful culti-
vation of Rhizobium outside the plant. Yet, he remarked, the three main
questions of the nodulation phenomenon remained: (i) how does the bac-
terium invade?, (ii) how is the nodule tissue induced? and (iii) how is
nitrogen fixation regulated in the nodule? Today, another 50 years on, the
answers are still not available.
The great contribution of genetics to the biochemical analysis of a wide
range of organisms is the result of the researcher's ability to use genetic
conditionality (i. e. mutant versus wild type) to provide a structure-function
relationship. This is more difficult to establish through biochemical or phy-
siological experiments alone. There is now a need to use the same
Genetics of Symbiotic Nodulation and Nitrogen Fixation 161
approach to study the mechanisms controlling symbiotic nitrogen fixation

and by doing so help to improve the legume symbiotic potential (see
Tudge, 1984). Dommergues (1978) stated: "Enhancing nitrogen fixation by
plant breeding is one of the most promising possibilities ... we would like
to reemphasise the need to develop research on plant genetics towards the
improvement of the relationship between plants and microorganisms". For
all of the above reasons we will focus here on areas of symbiotic nodu-
lation and nitrogen fixation relating specifically to plant genetics. Our
account will be somewhat speculative and personal, giving our work some
extra, and perhaps otherwise not necessary, emphasis.
II. A General Description of Legume Nodule Ontogeny
Not all legumes nodulate. However, most members of the Viciaceae and
Phaseoleae form nitrogen-fixing root nodules in symbiosis with the soil
bacterium Rhizobium. In general, temperate legumes (such as Pisum
sativum, Medicago sativa, and Trifolium species) develop indeterminate
nodules which are characterised by (i) a defined meristem during nodule
growth, (ii) an open vascular system connecting the root's vascular system
with the nodule meristem, (iii) vacuolated infected cells and (iv) aspar-
agine/glutamine as the major translocation products of nitrogen fixation.
These legumes tend to be infected by the bacteria belonging to the Rhi-
zobium genus (as compared to the newly recognised Bradyrhizobium genus
which contains bacterial species formerly incorporated in the Rhizobium
group; see Jordan, 1982). Tropical legumes of the Glycine, Vigna, Lupinus,
Macroptilium and Arachis genera develop nodules of the determinate type,
i. e. those in which the cell divisions responsible for the de novo synthesis
of bacteroid-containing plant tissue cease early in nodule ontogeny. This
means that mature, nitrogen-fixing nodules in contrast to the indeterminate
nodule type do not show meristematic regions. Figure 1 shows nodules on
roots of pea and soybean. Soybean nodules are spherical and pea nodules
cylindrical in shape. However, it is now clear that the plant genome con-
trols the nodule morphology. This is best exemplified by the Parasponia
symbiosis, in which the same bacterial strain induces two totally distinct
nodule types on two different plant hosts. Nodule sizes vary considerably
between species and even within the same plant. For example, on clover or
peas we have detected occasional nodule clusters of over 10 mm in
diameter. Such variation appears to be environmentally induced.
The mechanisms of infection of legumes by Rhizobium are varied and
mainly involve entry of the bacterium by an infection thread through the
extending tip region of a young root hair (Bauer, 1981; Dazzo and Gardiol,
1984; Rolfe and Shine, 1984) or through hydrolysis of middle lamellar
material of epidermal cells. Variations on these themes are found in stem
nodulation, as seen in Sesbania and Aeschynomene species (Legocki et al.,
1983; Tsien et al., 1983).
Fig. 1. Comparison of the nodule morphology of soybean and pea. The pea nodules
(left side) are of effective and ineffective type caused by a mixed infection of two
Rhizobium strains. The soybean nodules (right) are those formed on the taproot of a
supernodulating mutant. Note the distinct pattern of lenticel development
The invasion by Rhizobium induces sustained cortical cell divisions. Recent

analysis demonstrated that even prior to infection some cellular division
zones exist in legume root tissue. For example, Collins (1983) found cor-
tical "foci" located adjacent to the endodermis in infected and uninfected
white. clover roots. These were meristematic foci with up to 12 cells per
cluster, which coincided with the site of nodule initiation. Micro-
densitometry of nodules showed that a large proportion of nodule cells
have an increased DNA content. This led to the conclusion that the
invading Rhizobium induced polyploidisation in cortical tissues and that
this tissue was then stimulated to develop into the mature nodule. Ores-
shoff and Mohapatra (1981) reported studies with infected and uninfected
(nitrate grown) white clover plants, from which cortical cell protoplasts
were isolated by differential enzymatic digestion. These protoplasts were
Feulgen-stained and air-dried prior to Feulgen microdensitometry. Cortical
cells contained two distinct populations with DNA contents of 2C and 4C.
There was no evidence for mitotic divisions or S-phase nuclei. The 4C cells
thus were either tetraploid G 1 or diploid O 2 arrested. It was thus unnec-
essary to postulate the endoduplication induction by invading Rhizobium.
Appropriate target cells already existed. What is of significance is that

legume plants do show Grarrest in cortical cells. The arrest is achieved
through the translocation of a compound (7-methyl-nicotinate), which is
found in cotyledons of beans and peas, but not in those of two non-
legumes (Evans and Van't Hof, 1973). At present one can only speculate
about the involvement of this mechanism in the nodulation process.
Calvert et ai., (1984) used serial sections of infected soybean roots to
describe in detail the distribution of infection events. Many subepidermal
(hypodermal) divisions were not associated with infection threads or even
early infection events such as curling of the root hair. Many infection
threads were found to abort without eliciting a nodule. The hypodermal
infection foci were called "pseudoinfections" and could be induced only
by Bradyrhizobium (i. e. the normal soybean micro symbiont). The
induction did not require direct contact, as experimental separation of
plant and bacterium by a millipore membrane resulted in the same phe-
nomenon (Bauer, pers. comm.). Likewise it was seen that pseudoinfections
could be induced by an exogenous application of cytokinins such as
benzyladenine.
Figure 2 presents a model to explain the ontogeny of nodule devel-
opment in soybean. The initial stages encompass the findings of Calvert et
ai. (1984), while the latter ones are highly speculative and are designed to
include the observable structures of mature, nitrogen-fixing nodules. Of
importance to the theme of this chapter is the presence of plant structures
which also develop in other plant organs. Likewise it is suggested that Rhi-
zobium as the microsymbiont is restricted by the plant in its growth (mainly
through the plant-controlled milieu affecting oxygen and nutrient rela-
tions; Sinclair and Goudriaan, 1981; Witty et ai., 1985). Thus the plant is
actively involved in the prevention of a parasitic invasion by Rhizobium
(Werner et ai., 1984).
The nitrogen-fixing nodule is of benefit to the plant, because of its
ability to maintain bacteroids, which convert atmospheric nitrogen gas to
ammonia. Nitrogenase is made up of two major subunits, the iron protein
(also called nitrogenase reductase, coded for by the bacterial gene nifH;
molecular weight = 31000, Scott et ai., 1983) and the molybdenum-iron
protein (coded for by the nifD and nifK genes; molecular weight = 56000
each; Weinman et ai., 1984; see Brill (1980) for general references). The
nitrogenase complex requires ATP, electrons and redox potential to facil-
itate the reduction of nitrogen gas. The following reaction is thought to
best describe the stoichiometry:
N z + 10H+ + 16ATP+8e- -+ H z +2NH! + 16ADP+ 16 Pi
Hydrogen production by nitrogenase is an inescapable part of the reaction.

So far all nitrogenase systems produce hydrogen and therefore this process
is viewed as an energy loss to the bacterium and thus the plant. Several
Rhizobium strains possess a hydrogen uptake (hup) system, which "re-
cycles" hydrogen gas to form water and ATP (Schubert and Evans, 1976).
The hup system may be of value to the symbiosis, as H. J. Evans (pers.

comm.) has recently shown that hup+ strains of Bradyrhizobium japonicum
give up to 10 % better seed yields in symbiosis with soybean when com-
pared with hup- isolates. The ATP, electrons and redox potential are sup-
plied through the respiratory chain of the bacteroid. There exists evidence
that Rhizobium species possess several respiratory chains (Appleby, 1984).
The nitrogenase enzyme is extremely oxygen sensitive (Bergersen and
Turner, 1967, 1978; Mohapatra and Gresshoff, 1984). Exposure of isolated
bacteroids to atmospheric oxygen for just one minute eliminates most
nitrogenase activity through enzyme instability rather than transcriptional
control (Van den Bos et al., 1983). The regulation of transcription of
nitrogenase genes in Rhizobium bacteroids is different than that found in
free-living bacteria such as Klebsiella pneumoniae, although there may be
several similarities such as promoter sequences and nifA-like activation of
transcription (see Elmerich, 1984) for a review of the regulation of
Stage one: SlAge two:
St.g four: St.g. fly.: St.ge .Ia:
St.ge ven: St.ge eigh Khematk- ilde on

nitrogenase in free-living diazotrophs). Neither oxygen nor ammonia

repressed transcription of the nifHDK message of R. ieguminosarum
bacteroids (Van den Bos et ai., 1983). Likewise ntrC mutants of R. meliloti
were symbiotically active, although the development of the symbiosis was
delayed (Ausubel et ai., 1985).
Oxygen is needed for nitrogen fixation, as illustrated by studies with
isolated bacteroids under oxygen-free conditions (Bergersen and Turner,
1967). The legume host achieves a solution to the apparent paradox
through the development of regulatory and restrictive structures in the
nodule (Sinclair and Goudriaan, 1981; Ralston and Imsande, 1982). The
iron-containing leghemoglobin protein plays a key role in the facilitated
oxygen transport to the bacteroid (Appleby, 1984). Recent work by
Minchin and collaborators demonstrated the presence of a variable oxygen
diffusion barrier in several legume nodule types (Minchin et ai., 1983;
Sheehy et ai., 1983; Witty et ai., 1984). Carroll (1985) also provided evi-
dence, through a variety of experiments, which suggested the existence of
the oxygen diffusion barrier in soybean nodules. This barrier may be acti-
vated as soon as there is an alteration to the efficient functioning of the
nitrogenase system. It is even conceivable that a "feed-forward" control
Fig. 2. Model of the nodule ontogeny in the soybean-Bradyrhizobium symbiosis.

Stage one: the bacterium (arrow) infects an epidermal cell and elicits subepidermal
cell divisions. V = vascular bundle; P = pericycle (the site of lateral root for-
mation); E = endodermis; C = cortex; Ep = epidermis. Stage two: after pene-
tration of the infection thread of the root hair, cortical cells divide in advance of the
ramifying infection thread. Such hypodermal divisions can occur without infection
and are termed pseudoinfections. Stage three: increased cortical activity "signals"
pericycle tissue to initiate divisions. Stage four: progressing cortical activity results
in a visible lump. Further pericycle divisions distort and eventually rupture the
endodermis. Stage five: bacteria leave the infection thread and start bacteroid
development leading towards nitrogenase activity. Pericycle tissues push through
the endodermis and proliferate towards invaded (shaded) zone. Nodule expansion
may break through the epidermis (this is not shown on the diagram). Stage six: con-
tinued development of pericycle and possibly cortical cell material to develop vas-
cular traces surrounding the infected zone. Tracheids at first develop in parallel to
the root axis, often in unconnected cell clusters, but then "grow" together to form
perpendicular vascular strands (indicated by the spiral tissue). The early boundary
cell layer around the infected zone is noticeable as is the formation of an endo-
dermis-like layer in the nodule cortex. Stage seven: the infected zone is fully
nitrogen fixing and is characterised by infected and uninfected cells. The latter type
harbours the enzymes related to ureide metabolism. Vascular bundles are con-
nected and surround the infected tissue. Scleroid (thick-walled) cells develop in a
nearly continuous layer in the nodule cortex. Gaps in this layer are associated with
the position of the vascular bundles, which also govern the location of the lenticels
(LC). Stage eight: as stage seven, but indicating some possible cell lineages. P =
pericycle; PC = plant cortex; LC = lenticel; SC = scleroid cell; IZ = infected
zone; VB = vascular bundle; EP = epidermis (casparian strip); NOC = nodule
outer cortex; NIC = nodule inner cortex; PDT = peri cycle derived tissue; BCL =
boundary cell layer.
loop is in action, as the plant itself does not appear to have a chance to
monitor the altered nitrogen output of the bacteroid within the one to two
minutes that are taken to facilitate the barrier's action. Anatomically this
barrier resides in the cortex of the nodule as confirmed by direct measure-
ments using microelectrodes (Tjepkema and Yocum, 1974; Witty et al.,
1985). The cortex contains a layer of scleroid-type cells which is not con-
tinuous and shows gaps above the position of the vascular traces. Lenticels,
being parenchyma-like outgrowths on the nodule surface, also follow the
vascular traces (see Fig. 1 for well-developed lenticels on the surface of the
soybean nodules). Pankhurst and Sprent (1975 a, b) found that desiccation
or waterlogging inhibited nitrogenase activity of nodules. This loss of
activity was correlated with the anatomical collapse of the lenticel
structure. Similarly, the same authors found that elevated oxygen concen-
trations reversed the waterlogging-induced decrease in nitrogenase. Dark-
induced and nitrate-induced losses of nodule nitrogenase activity were also
recovered after nodules were incubated at elevated oxygen levels (Carroll,
1985).
In general it is arguable that legume nodules regulate bacteroid
nitrogenase activity not through the controlled supply of carbohydrates
(say malate or succinate), but through the provision of oxygen, which is
needed for oxidative phosphorylation. Additional barriers to oxygen
transport into the nodule interior are (i) the lack of airspaces in the inner
cortex (Bergersen and Goodchild, 1973), (ii) the bacteroid-boundary cell
layer (see Fig. 2) and (iii) plant respiration and cytoplasmic oxidases which
lower the oxygen concentration close to the bacteroid surface. Bacteroids
scavenge low amounts of free oxygen (being about 1-10 micromolar in
most legumes) through the interaction of leghemoglobin and specific oxi-
dases which accept oxygen from the hemoglobin molecule. This protein is
an example of a molecular symbiosis. The globin moiety is a product of
plant genes which are directly related to globin genes found in animals
(Appleby, 1985; Verma et al., 1985). The heme molecule is the presumed
product of the bacterium and is excreted into the plant cytoplasm during
bacteroid development. However, there is no evidence for heme export into
the peribacteroid space or for heme transporter proteins in the bacterial
and peribacteroid membranes. Part of the argument for the bacteroid
origin of heme is the observation that heme biosynthesis was regulated by
micro aerobic conditions similar to those expected in the developing nodule
(Nadler and Avissar, 1977; Keithley and Nadler, 1983). Furthermore, there
is an apparent correlation between bacteroid content and heme content in
soybean nodules (D. Day, pers. comm.) and genetic data from Rhizobium
meliloti mutants, which lack the first enzyme of the heme biosynthetic
pathway (being delta-aminolevulinic acid (ALA) synthetase), provided evi-
dence for the bacterial origin of the heme moiety. The mutants produced
white nodules while their revertants gave red nodules (Leong et al., 1982;
Ditta et al., 1983). Such an argument is,.however, not sufficient, as one
finds that, for instance, leucine-requiring auxotrophs of R. trifolii also form
white nodules and that their revertants are symbiotically effective (Bassam
and Gresshoff, 1986). Yet bacterial leucine biosynthesis is not directly

essential for leghemoglobin biosynthesis. The classical assumptions of this
process were questioned recently through the isolation of an ALA-syn-
thetase-deficient mutant of Bradyrhizobium japonicum which still nodulated
and gave leghemoglobin-containing nodules (Guerinot and Chelm, 1985).
Whether ALA was crossfed from the plant to the bacteroid for further pro-
cessing, or whether an alternative pathway existed in the bacterium, or
whether the plant after all supplied the heme, is not clear.
Robertson et al. (1984) using immunogold labelling localised leghemo-
globin in the cytoplasm and not in the peribacteroid space of pea and lupin
nodules. Verma et al. (1985) obtained similar results with soybean nodules.
How oxygen is transported across the peribacteroid space is not clear.
Simple diffusion could suffice, especially since the volume of the peri-
bacteroid vesicle may be enlarged in electron-micrographs due to fixation
artifacts (G. D. Price, pers. comm.). In relation to this point and the next, it
is clear that there are various aspects of symbiotic nitrogen fixation where
the final analysis of the mechanism has not been completed. In other
words, gray areas remain in mechanisms which could have been studied by
existing technologies of biochemistry and microbiology. For example, the
succinate-uptake mutants of Ronson (see next paragraph) still induce
leghemoglobin-containing nodules. However, bacterial ALA synthetase
uses glycine and succinate as substrates. So where does the succinate-
uptake mutant, which is really struggling to get enough dicarboxylic acids
to maintain nitrogenase, obtain the succinate to permit an excessive
excretion of heme to the plant host? It is these shortcomings or gaps in the
knowledge of the basic biology of nitrogen fixation that hopefully can be
closed by the application of biochemical genetics of the plant and the bac-
terium.
Genetic studies by Ronson et al. (1981, 1984) substantiated the bio-
chemical and physiological findings of the importance of dicarboxylic acid
transport to the nitrogen-fixing bacteroid. Thus bacterial mutants, which
lacked the succinate-malate-fumarate uptake system were capable of nodu-
lation and bacteroid development, but not of nitrogen fixation. The dctA
gene has been cloned and sequenced. Hydropathic analysis of the amino-
acid sequence showed 12 transmembrane helices consistent with a mem-
brane-associated molecule. Analysis of the carbon requirement of isolated
bacteroids also implicated the critical role of succinate-malate. Often it was
claimed that succinate increased nitrogenase activity but this was a res-
trictive interpretation, as it depended on the external oxygen concen-
tration. Investigations with isolated bacteroids from soybean and Para-
sponia rigida (a non-legume nodulated effectively by certain Bradyrhi-
zobium strains) showed that maximal nitrogenase activity did not change
through the addition of succinate, but that instead the oxygen sensitivity
profile was altered. This meant that in the presence of higher oxygen levels
during incubation, succinate increased detectable activities (McNeil et aI.,
1984; Carroll, 1985; Sandeman and Gresshoff, 1985). McNeil et al. (1984)
showed that optimum nitrogenase activities over the first 100 min. at six
levels of oxygen between 0.1 % and 1.5 % in the gas phase did not differ
between the additions of 50 mM sucrose, 50 mM glucose or 50 mM suc-
cinate. These optimum activities were identical to the "no carbon-source"
control, suggesting that soybean bacteroids had large endogenous supplies
of available carbon source. If bacteroids were starved of carbon in planta
by preincubation of the plant in the dark for 2.5 days and assayed after
anaerobic isolation for initial nitrogenase activity, then the addition of 50
mM pyruvate or succinate (even as low as 1.5 mM) resulted in substantial
increases of nitrogenase activity. Dicarboxylic acids other than succinate
may be of physiological importance. For example, the carbon compound
most important in the support of nitrogen fixation may be malate, which is
produced by phospho-enol-pyruvate carboxylase (PEPC) in the nodule
cytoplasm without "taxing" the already "stressed" nodule mitochondria.
PEPC is a major enzyme found in nodule cytosols (Gadal, 1983). A more
speculative suggestion was made by Kahn et al. (1985), who proposed that
glutamate may be a key compound transported into the Rhizobium cell. It
may be of value to test this hypothesis with dct mutants or with isolated
bacteroid systems as described above.
Ammonia produced by nitrogenase diffuses out of the bacteroid. There
is no direct evidence for an ammonia export system under bacteroid condi-
tions, as the plant cytoplasm contains high amounts of plant-derived gluta-
mine synthase which assimilates the ammonia with a low Km to form gluta-
mine, which in turn serves as a translocation product or is used for trans-
amination reactions (Streeter, 1977; Duke and Henson, 1985). In tropical
legumes the major translocation products are the ureides (Schubert, 1981;
Hanks et al., 1983). The level of ureides in the xylem, especially if
expressed as relative ureides, has been taken as a measure of the amount of
nitrogen fixed (Herridge, 1982). McNeil and LaRue (1984) demonstrated
that ureides were also produced by nitrate-grown and non-nodulated
soybean plants. Additionally it should be considered that the determi-
nation of the relative ureide level assumes that nitrate uptake is constant
between comparative samplings. This may be so within the same cultivar
monitored at the same developmental stage, but may differ significantly
between different cultivars. The conclusion is that we do not even know
how to measure nitrogen fixation in the field accurately!
III. The Parasponia-Bradyrhizobium Symbiosis
As stated above, Rhizobium and Bradyrhizobium species form a variety of

nodule types on a range of legumes. This was taken in the past as a basic
characteristic of Rhizobium. One notable exception to the legume speci-
ficity of Rhizobium is the nodulation of plants within the Parasponia genus
by many Bradyrhizobium isolates. This was first discovered by Trinick
(1973), although the plant species was misidentified as the related species
Trema. The Trema genus has not yet been reported to nodulate, although it
belongs to the Ulmaceae, of which several species are nodulated by the bac-
terium Frankia (Lalonde, 1979). Since then numerous studies have utilised
the Parasponia system and progress has been reviewed by Gresshoff et al.
(1984). A few distinct features will be highlighted here. The association is
as efficient as the legume symbiosis in terms of nitrogen fixation, although
nodule number per plant in Parasponia is increased, and the amount of
infected tissue per nodule is decreased when compared to the legume (such
as siratro, Macroptilium atropurpureum) control grown under identical con-
ditions with the identical bacterial inoculant. The Parasponia nodule seems
similar to a modified lateral root as it has a central vascular cylinder.
Nodule growth starts in the pericycle and not the cortex. The infected zone
contains infected and uninfected cells with the hemoglobin being localised
in the cytoplasm of the infected cell. The amino-acid sequence of the P.
anderson;; hemoglobin shows very close homology to that of the soybean
Lb c hemoglobin, suggesting a common evolutionary origin (Appleby et al.,
1983; Kortt et al., 1985). The preliminary Parasponia hemoglobin
sequences referred to by Gresshoff et al. (1984) appear to be artifacts and
erroneous. Bacteria in the Parasponia nodules apparently fix nitrogen,
while still contained in the modified infection thread. This structure has
been termed the fixation thread (Price et al., 1984).
The infected zone is surrounded by a layer of cells which show
extensive tannin accumulation, especially in the vacuole (Price et al., 1984).
This deposition of phenolic compounds may be related to the recognition
by the plant of the micro symbiont as a potential pathogen (compare to
later discussion of the gene-for-gene concept).
Isolated bacteroids of Bradyrhizobium strain ANU289 obtained from
nodules of P. rigida showed similar oxygen sensitivities as siratro-derived
bacteroids of the same strain (Sandeman and Gresshoff, 1985). In vitro
derepression studies of strain ANU289 indicated the further similarities to
other Bradyrhizobium strains (Mohapatra and Gresshoff, 1984). This simi-
larity was further confirmed by molecular analysis of the nitrogenase
genes, which share sequences as well as the genomic arrangement with
other bacteria which infect soybean or peanut but not Parasponia (Scott et
al., 1983; Weinman et al., 1984). This non-legume symbiosis demonstrates
clearly that one specific Bradyrhizobium strain can be flexible enough to
invade a root, induce a morphologically different nodule, and maintain a
functional symbiosis in two distinctly different plant hosts. It also suggests
that the two hosts are capable of supplying a relatively similar biochemical
and physiological niche. Hence it is possible that many other non-legume
plants are capable of infection and nodulation by Rhizobium, although
these may not have been recognised as yet. The taxonomists of the last two
centuries have seldom included root characteristics in their botanical
writings. The question arises now as to the nature of the plant genes which
facilitate infection and nodule formation in this symbiotic system. Are they
similar to legume genes? Do other plants possess similar genes? How does
the bacterium nodulate the two hosts? Are different bacterial genes
employed? What is the function of symbiotically activated genes in non-
symbiotic tissue and/or species?
IV. Biochemical and Molecular Analysis of Plant Functions
There are numerous plant gene products involved in nodulation and

nitrogen fixation. Many should also have "normal" plant functions, which
are utilised in this alternative developmental pathway. Others will be more
or less specific. Recent molecular analyses have focussed on the presence
of nodule-specific proteins called nodulins (Bisseling et al., 1983, 1985;
Fuller and Verma, 1984; Fuller et al., 1983; Miflin and Cullimore, 1984;
Verma and Nadler, 1984; Govers et al., 1985; Verma et al., 1985). The
definition of the term nodulin may also have to include nodule-amplified
proteins. At present the concept also seems to focus on the root as a control
tissue. This is not a necessity, as some symbiotic functions may utilise
genes which are only expressed in other tissues at another time in the
plant's ontogeny.
The major approach has been the production of nodule-specific
antisera which were used to visualise peptides on two-dimensional gels.
Antisera against nodule proteins were titrated with excess root proteins,
thereby precipitating the "common" antibodies. The remaining uncorn-
plexed antiserum contained antibodies which were predominantly specific
for nodule proteins. If used in a Western blot, these sera would recognise
the nodule-specific peptides and this was visualised by a second reaction in
which the now-complexed antibody was recognised by a second antibody
which carried a means of visualisation. Usually this was a radioactive
iodine tag or a covalently-linked enzyme such as peroxidase. The presence
of the antibody-protein interaction can thus be recognised by either autora-
diography or colorimetric methods. In a similar fashion, it was possible to
prepare cDNA from nodule RNA and to use such molecules in the differ-
ential screening experiments. To do so, cDNA libraries from root and
nodule tissues of similar age were screened for differential hybridisation
against nodule-derived cDNA probes. Relevant nodule-specific clones
were then analysed, rechecked for specificity, sequenced and used to
isolate genomic sequences of homology. This procedure allowed the iso-
lation of three characterised nodulin genes. These are the leghemoglobin
family (Lee et al., 1983), glutamine synthetase (Cullimore et al., 1983) and
nodulin 35 (being the enzyme uricase, which is essential in the ureide bio-
synthetic pathway; Legocki and Verma, 1979; Bergmann et at., 1983).
Major advances have been achieved in the pea and soybean system (for
reviews see Bisseling et al., 1983, 1985, as well as Verma et al., 1985). In
pea, Bisseling and collaborators found several (up to eight) nodulins which
were coordinately expressed with leghemoglobin. This occurred whether
the nodule was induced by a bacterium capable or incapable of nitrogen
fixation. Similarly the Rhizobium chromosomal (as compared to plasmid)
replicon was required for complete nodulin expression, as Agrobacterium
tumefaciens transconjugants showed only one nodulin. Similar results were
obtained in soybean and in alfalfa (Medicago sativa). The composite data
of Bisseling et al. (1985), Lang-Unnasch and Ausubel (1985) and Moha-
patra and Piihler (1986) argued for a cascade-type of activation during
nodule development. A key event may be the success of the invasion within
the first 24 hours after infection. This apparent escape from the host
defences is then followed by the plant regulation of further infection and
nodule proliferation (autoregulation) and finally the nodule control over
nitrogen fixation through the control of oxygen diffusion to the encased
bacteroid. Thus the plant successfully maintains control over the infection
through a series of coordinated developmental steps.
In our laboratory we tried to test whether this coordination of sym-
biotic steps is maintained if the symbiosis is restricted or inhibited by
external factors. Schuller et al. (1986) applied 10 mM nitrate to established
symbiotic soybean plants and followed a number of symbiotic parameters.
They found that nitrate inhibited nitrogenase activity (up to 50 % inhibition
after 2 days), but that all nodule parameters such as leghemoglobin, glut-
amine synthetase, ureide export, uricase, xanthine dehydrogenase, and iso-
lated bacteroid nitrogenase activity were maintained at control levels, until
general nodule senescence set in after about 14 days, when most relevant
peptides and their activity disappeared. It is thus likely that the symbioti-
cally important plant functions, once induced in a coordinate way, are
maintained and that no further gene regulation is possible other than
through general abscission of the organ. These findings fit to those related
to cascade-type inductions and eventually may tell us more about determi-
nation and differentiation in plants.
After the significant findings of bacterial genetics leading to the defini-
tion of the nodulation genes (Schofield et al., 1983, 1984; Kondorosi et al.,
1984; Schmidt et al., 1984) and the recognition of plant exudates (Hal-
verson and Stacey, 1985; Olson et al., 1985; Mulligan and Long, 1985) and
lectin binding features, several researchers have turned their attention to
the plant's direct contribution in the symbiosis. Although the bacterium
harbours the causative factors (nod genes, nifgenes and perhaps heme bio-
synthesis genes), the plant has retained control over the key processes.
Such considerations are progressively becoming important for those who
desire to improve commercially the legume-Rhizobium symbiosis (see also
a general article on this subject by Barton and Brill, 1983; a comprehensive
review dealing with the potential of Rhizobium improvement was prepared
by Hodgson and Stacey, 1986).
v. Gene-for-Gene Aspects of Nodulation

Pathogenic interactions between plants and microbes are of two main types
- either successful invasion and multiplication of the pathogen at the
expense of the plant, or successful resistance of the plant at the expense of
the pathogen, preventing its invasion and increase. Situated in neither of
these camps is the legume/Rhizobium symbiosis, in which both plant and
bacterium benefit from their association, but which nevertheless has many
features in common with the usual plant/pathogen interaction (Vance and
Johnson, 1983). These features include the attachment of bacteria to the
host plant, their successful penetration, the response of the host to this
invasion, a change in plant metabolism and obvious morphological
changes as a result of the infection. .
As in any form of interaction between two organisms there must be a
genetically controlled recognition sequence, which determines whether or
not the symbiosis or pathogen attack will be successful, or elicit a negative
interaction such as a hypersensitive response, phytoalexin accumulation or
similar (Albersheim and Anderson-Prouty, 1975; Daly, 1984; Sequira,
1984). This interaction is a function of genes present in both bacterium and
plant. Whereas considerable information is available on the genes which
condition successful nodulation in Rhizobium, much less, as stated in the
introduction, is known about the plant genes. It has been estimated that at
least 10 plant genes are involved in nodule development alone (Holl and
LaRue, 1976). A similar number may also be controlling the infection
process as well as the functioning of the nitrogen fixation processes,
including ammonia assimilation, carbohydrate supply and oxygen dif-
fusion. Whatever this number, it is large and thus gives potential for
genetic variation within the plant. This in turn has phylogenetically
resulted in the specificity of infection and nitrogenase expression seen in
efficient symbioses. The exact nature and mechanics of this specificity, and
the genetic components of its control are not well understood, particularly
in relation to the plant. In other successful pathogen attacks a gene-
for-gene relationship has been demonstrated (Flor, 1955, 1971; Day, 1974),
which describes the response of the host plant to various gene products
from the pathogen, but does not, in most cases, explain the way in which a
host plant recognises infective or non-infective pathogens. With the Rhi-
zobium-legume symbiosis involving fast-growing Rhizobium strains the
genes coding for the ability to recognise a host plant are located on large
indigenous plasmids (Djordjevic et ai., 1983, and many others). These sym-
biosis (Sym) plasmids can be removed (cured) from the strain, rendering it
unable to interact with the legume host. When transferred to a different
Rhizobium species, the presence of the Sym plasmid allows initiation of
nodulation but not nitrogen fixation in the new host (Rolfe and Shine,
1984), thus indicating that at this point specificity is controlled by genes on
the chromosomal rather than the Sym plasmid replicon. In Bradyrhizobium
the symbiotic genes are not located on plasmids (Fischer and Hennecke,
1984; Masterson et ai., 1985). The use of large deletions spanning the sym-
biotic region, and replacement analysis, however, have aided complex
genetic analysis, so that the absence of symbiotic plasmids is no longer a
problem to further genetic investigation of Bradyrhizobium.
Whatever systems are involved, they must include an interchange of
signals which enable more specialised pathogens to avoid being identified
by the host defence systems as foreign, and allow them to continue their
development within the plant tissue. An incompatible combination of
pathogen or host results in a reaction which effectively prevents any
further pathogen development. That plants can respond to microorganisms
has been known from the beginning of this century; the term "hypersen-
sitive response" was coined for the reaction, in which cells in a resistant
host adjacent to the infection site rapidly become discoloured, granular
and necrotic (Klement, 1982). There are many examples of this type of
disease resistance to be found in cereals against rusts and mildews, in
potato against Phytophthera infestans, and in other plant species against
bacteria, viruses and nematodes (Misaghi, 1982). In all cases the response
is similar. In comparisons of the mode of host response to virulent and
avirulent pathogens, no difference was detected in the way in which the
epidermis was initially penetrated. After infection, however, resistant plant
varieties showed a loss of turgour, browning of localised tissues, and cel-
lular death. These changes in the cell were associated with loss of permea-
bility of the plasma membrane, increased respiration, accumulation and
oxidation of phenolic compounds and the production of phytoalexins
(Keen and Kennedy, 1974; Heath, 1980). The infected cell and a few sur-
rounding ones died, and the invader was engulfed by necrotic tissue which
blocked successful proliferation of the pathogen. This mechanism appears
commonly in a range of plants infected by a range of microorganisms, and
thus may indicate a common resistance mechanism in plants to perceived
pathogens (Heath, 1981). Halverson and Stacey (1986) reviewed the general
area of plant-microbe interaction and discussed in detail aspects of the
molecular signalling between host and potential pathogen. (See also
Chapter 8 in this volume).
Recent studies have concentrated on the host-pathogen interaction in
early stages of infection, with particular reference to host plant responses
(Fraser, 1982; Misaghi, 1982). Avirulent strains of Pseudomonas solana-
cearum which attached only weakly to cell walls of tobacco and were
agglutinated by a hydroxyproline-rich glycoprotein extracted from potato
tubers, lacked the O-antigen oligosaccharide of the lipopolysaccharide
(Keen and Williams, 1971; Billing, 1982). Electron-microscopic analysis
showed that the pathogen, when infiltrated into tobacco leaves, was closely
attached to plant cell walls and at a later stage became enveloped in a
fibrillar material. This did not happen with virulent strains of the pathogen,
which multiplied in the intercellular spaces. Clearly the plant recognised
the avirulent strain but not the virulent one.
With the P. solanacearum system the recognition is known to be asso-
ciated with lectins in the plant, which bind to the lipopolysaccharide (LPS)
of the bacterial cell wall. Purified LPS binds to plant walls in the same way
as do avirulent bacteria. LPS of both virulent and avirulent strains behave
in a similar way. The only difference is that virulent bacteria produce extra
slime (mainly in the form of exopolysaccharides, EPS). This additional
external "dressing" may prevent the recognition of the active LPS factors
by the plant lectin and the concomitant hypersensitive response (Billing,
1982).
The legume-Rhizobium symbiosis, as stated above, requires a number of
precise interactions and different responses from the plant to develop the
functioning nodule. Interaction is probable prior to attachment (cf. pseu-
doinfections in soybean), during attachment, infection thread growth and
penetration into neighbouring cells, induction of cell proliferation to

initiate nodule growth, bacteroid release from the infection thread, and
during the bacteroid development-bacteroid maintenance phase. Rhi-
zobium mutants have been isolated which have altered EPS production.
Their symbiotic phenotypes vary, but in general one finds less efficient
symbioses, less nodules, and stronger rejection of the bacterium (Johansen
et al., 1984; Muller et al., 1985). The EPS of Rhizobium may be important as
a protective barrier when inside the root environment, but may also be
necessary to elicit the correct response from the plant to complete the nod-
ulation sequence. Mutants altered in their EPS biosynthesis in general are
recognised as pathogens and cause a hypersensitive response thereby pre-
venting further growth. For example, the nodules induced by the R. meliloti
EPS-deficient mutants decribed by Muller et al. (1985) were empty of bac-
teria. The nodules were tumour-like and showed obvious signs of initial
bacterial infection as the nodule surface showed small infection pockets.
These were clearly enclosed and aborted infections. The cell wall sur-
rounding the infection pocket showed thickenings and clear secondary cell
wall deposition. Of interest is that the correct set of nodulins were still
induced, including hemoglobin (Mohapatra and PUhler, 1986), suggesting
that the nodulation induction signal was only required in the early stages
of the interaction.
Analysis of soybean nodules formed by an ineffective Bradyrhizobium
strain (61-A-24) revealed an increase of the phytoalexin glyceollin, which
occurred in the root or nodules formed by wild-type bacteria (Werner et al.,
1985). The ineffective nodules lacked hemoglobin and had an unstable
peribacteroid membrane. Nodules were numerous and small compared to
wild type. Another jix- mutant lacked these symbiotic phenotypes. The
data suggest that in some Rhizobium mutants the latter stages of symbiotic
communication are defective, resulting in the host's recognition of the
invader and the accompanying phytoalexin effect.
VI. Existing Plant Variation in Symbiotic Nitrogen Fixation
It has long been recognised that the plant host genotype influenced the
major symbiotic characteristics (Nutman, 1946, 1949, 1953; Holl, 1983).
This was despite the failure of plant breeders and agronomists to recognise
the importance of nodulation and nitrogen fixation to plant productivity.
For example, in a recent paper Nelson and Bernard (1984) investigated the
production and performance of hybrid soybean. They described heterotic
interaction for characters such as yield, maturity date, lodging, height,
harvest index, percentage oil and protein, seed quality and weight, but not
symbiotic characters.
A major input into the recognition of plant cultivar effects came from
multi strain - multicultivar interaction trials (Pacovsky et al., 1984).
Certain cultivars do better with certain Rhizobium strains. For example,
Rennie and Kemp (1984) compared the nitrogen fixation abilities of two
bean (Phaseolus vulgaris) cultivars by lsN isotope dilution technique. They

found that cultivar Aurora had superior nitrogenase activity at 12 C com-
pared to cultivar Kentwood (average 119 kg N/ha versus 84 kg N/ha at 10
kg N/ha supplied nitrogen). This difference was environmentally in-
fluenced, so that culture at 15C negated the superiority of Aurora. The
essential finding was that Aurora was superior because of a longer vege-
tative phase supporting nitrogen fixation. Similar studies with soybean
(Rennie and Dubetz, 1984) also demonstrated differences (e. g. cultivar
Maple Amber fixed 91 kg N Iha while cultivar Maple Presto fixed 60 kg
N/ha). The strong environmental influence may be explained by the
apparent multigenic nature of this type of variation involving complex
gene and strain interactions.
Differential nodulation patterns and nitrogen fixation rates were seen
in many legume species. Harper and Gibson (1984) compared the nitrate
sensitivity of soybean, lupin, chickpea, siratro, pea, subterranean clover,
lab lab bean and barrel medic in hydroponic solution. Clear differences
emerged, with soybean, lab lab bean and barrel medic being more sensitive
to nitrate in terms of nodulation than the remainder of the tested species.
Similar variation in nitrate sensitivity has been reported by Herridge and
Betts (1985), who tested over 480 soybean cultivars and found continuous
variation for symbiotic parameters such as nodulation index and ureide
content. Their work claimed that two cultivars of Korean origin had
increased nodulation and apparent nitrogen fixation on nitrate-containing
soils, when compared to a control cultivar Bragg. Whether it is possible to
use such variation in breeding programmes is not known at the moment
and will depend on the genetic nature of the variation.
Skot (1983) found cultivar differences in peas, so that at 63 days after
planting the plant genotype accounted for 75 % of the variation in dry
weight, while the Rhizobium inoculum accounted for 63 % of the variation
in nitrogen content and 70 % of the variation in dry weight/total N ratio.
Plant control over nitrogen fixation was demonstrated by Thomas et al.
(1983), who inoculated soybean variety Wells with two different Bradyrhi-
zobium japonicum mutants which had an enhanced ability for nitrogen
fixation. Although improvements were seen by three weeks (most likely
caused by increased early nodulation and perhaps competition), by five
weeks there was no difference in N content of the xylem sap, acetylene
reduction, leaf allantoinase activity and shoot ureide content.
Sen and Weaver (1984) studied the differential nitrogen fixation rates
observed when the same cowpea Bradyrhizobium strain was tested on
peanut and cowpea. Bacteroids isolated from both hosts had similar rates
of oxygen consumption and acetylene reduction. Although the number of
bacteroids per unit nodule volume was higher in cowpea than in peanut the
nitrogenase activity as measured on whole plants showed that peanut was
more active. This suggested a partial oxygen deficiency within the more
"crowded" cowpea nodule and the conclusion was supported by the fact
that increased partial pressures of oxygen increased the observable
nitrogenase activity in cowpea plants.
Phillips and associates showed that the plant genotype affected

hydrogen evolution by R. leguminosarum induced nodules on pea (Bedmar
and Phillips, 1983; Bedmar et aI., 1983). A shoot factor in pea regulated the
degree of hydrogen evolution (Bedmar and Phillips, 1984). Similarly,
alfalfa cultivars were selected from existing material to produce lines with
consistently higher nitrogen fixation rates. This increase was reflected in
improved fitness in the absence and presence of exogenously supplied
ammonium nitrate. Thus, performance of selected line HP32 exceeded that
of HP for dry weight and total reduced nitrogen by 42 % and 45 % in
ammonium nitrate-dependent plants and by 69 % and 81 % in plants
dependent on nitrogen fixation alone (Phillips et al., 1985). In field condi-
tions, such differences were reduced to forage matter increases of 10 % and
total forage nitrogen increases of about 22 %. Whilst agronomically signif-
icant, and again demonstrating the importance of the plant genotype in the
joint symbiotic phenotype, the application of existing variability to bio-
chemical analysis was hindered by the general inability to explain the
causes for the improved line. This again goes back to the possible multi-
genic nature of such phenotypic differences, which will result in mUltiple
biochemical changes.
Leghemoglobins of Medicago sativa, Lathyrus, Lupinus polyphyllus and
subterranean clover were evaluated by cellulose acetate electrophoresis
(Holl et al., 1983). Although different profiles were observed for different
species, no qualitative differences were found in almost 1000 plants of dif-
ferent cultivars of alfalfa. Only one plant of subterranean clover line
PS-1003 failed to produce the major component of LbI, even if inoculated
with different Rhizobium strains. This distribution was not affected by the
nodule age.
The consistency of the Lb profile within one species may stem from the
restricted gene pool found in many commercial crop species. Smartt (1984)
reviewed the gene pools of grain legumes and concluded that utilisation of
the plant genetic resources, unlike many other of our resources, does not
bring about exhaustion, when they have been collected and conserved.
Delannay et al. (1983) found that the northern and southern gene pools of
soybean have maintained remarkable qualitative uniformity since 1951,
showing that plant breeding techniques had used the same material re-
peatedly.
Bowen and Kennedy (1961) studied heritable variation in Centrosema
pubescens and found that commercially available seeds contained stable
sparsely nodulating and profusely nodulating lines. With most Rhizobium
strains the sparsely nodulating lines were inferior in symbiotic effec-
tiveness.
Jones (1962) described considerable genetic variation in nodulation
parameters in white clover. Total nodule weight per plant was highly corre-
lated with eventual harvest weight of the plant. Even other nodule-related
parameters like bacteroid viability were strongly dependent on the host
plant (Sutton and Paterson, 1983).
VII. Existing Single Locus Variation for Nodulation-Nitrogen Fixation
A relatively large number of naturally occurring plant mutants have been

found which show alterations in their symbiotic phenotype. These differ
from the varietal differences described above in that their genetic basis can
be elucidated. In many cases the mutant phenotype was caused by a single,
identifiable, genetic alteration.
Five nodulation-nitrogen fixation mutants of soybean were described
prior to the present studies using induced mutagenesis (Carroll et ai.,
1985 a, b). The mutants are labelled rjh Rj2, Rj3, Rj4, and Rjs. Only one of
these (mutant rjl) was studied in any detail (Williams and Lynch, 1954;
Clark, 1957; Tanner and Anderson, 1963; Weber, 1966; Devine and Weber,
1977). It was restricted in its nodulation of soybean by Bradyrhizobium and
was controlled by a recessive allele. All others resulted in ineffective (non-
nitrogen-fixing) nodules, were dominant and showed strong strain depen-
dence. For example, mutant Rjs was recently found in cultivar Evans,
which was homozygous dominant for Rjs and gave nod+ andfix- symbioses
with strain USDA205. Cultivar Peking, in contrast, was rjsrjs (Devine,
1985). Mutations Rj2 and Rj3 gave small white nodules and were inde-
pendent dominants with particular strains belonging to the CI and 122
serogroups of Bradyrhizobium (Caldwell, 1966; Vest, 1970). Mutant Rj4,
which gave normal nodulation in early phases of development, then led to
totally ineffective symbioses with strain USDA61 (Vest and Caldwell,
1972).
Mutant rjh which was characterised by an inability to form nodules
with most Bradyrhizobium strains at normal inoculation densities, mapped
on linkage group 11 of soybean, being 40 map units distant from the F
locus, which controls fasciated stem formation in soybean (Devine et al.,
1983). The rjl mutation was bred into many commercial cultivars and serves
as a convenient non-nodulation control for nitrogen balance experiments,
and in the determination of nitrogen fixation in the field using relative
abundance and isotope dilution techniques. The mutant was found to be
conditional in its phenotype as Clark (1957) showed that strains such as
USDA76 were able to nodulate the mutant soybean sparsely. At first it was
thought that the ability to produce the compound rhizobotoxin was corre-
lated with the ability to suppress the mutant phenotype. However, recent
work by La Favre and Eaglesham (1984) disproved this. The same authors
also showed that increased inoculation densities of Bradyrhizobium
increased the degree of nodulation. For example, strain USDA76 at 2 x
1011 cells per pot gave about 48 nodules per plant (compared to 146 on
RjlRjl) and about 1.5 nodules per plant at 109 cells per pot on rj1rjl' The
correlation with rhizobotoxin production broke down as strains SM35 and
IRj2133b (both being unable to produce the toxin) gave 30 and 21 nodules
per plant respectively at high inoculation densities. Other strains like 1-110
nodulated poorly (about 320 nodules per plant on wildtype compared to
about 2.5 nodules per plant on the mutant). Other strains like IRj2147 and
IRj2179, despite their ability to produce rhizobotoxin, gave no nodulation
even at high inoculation densities. The apparent blockage in mutant rj, is

very early in the nodulation process, as no infections were observed in
normal nodulation trials. The genotype of the root controlled the pheno-
type and it was claimed that a diffusable rhizosphere factor (acting as an
inhibitor) was involved (Elkan, 1961), although this has been disputed
(Eskew and Schrader, 1977). Despite the presence of this genetic material,
little biochemical analysis was carried out on the rj, system.
Other naturally occurring mutants were decribed in Trifolium pratense
(5 genes, Nutman, 1969), Trifolium incarnatum (controlling a jix- pheno-
type, Smith and Knight, 1984), Medicago sativa (two clones with early se-
nescence, two other clones with tumour-like nodules with infection threads
or bacteroids, one with nodulation resistance, Viands et aI., 1979; Peterson
and Barnes, 1981), Arachis hypogaea (non-nodulation, Gorbet and Burton,
1979; Nambiar et al., 1983), and Pisum sativum (Gelin and Blixt, 1964; Lie,
1971; Holl, 1975; Ohlendorf, 1983 a, b). The general nature of these
mutants was described in the above references and in reviews by LaRue et
al. (1985) and Verma and Nadler (1984). A general conclusion that can be
drawn from the entire spectrum of non-nodulating and non-nitrogen-fixing
plant mutants is that (a) they all exhibit strain specificity, (b) either dom-
inant or recessive types exist, (c) they exist in a large range of legumes, (d)
they were not numerous enough to have allowed an extensive biochemical
analysis and (e) they seldom were the product of a directed search but
rather appeared through serendipity.
VIII. Induced Mutation in Symbiotic Characters
Major advances in biology have stemmed from the tight coupling of bio-
chemical and genetic approaches. The studies on Neurospora by Beadle
and Tatum, the Escherichia coli lactose operon concept as initially pro-
posed by Jacob and Monod, the analysis of development in bacteriophages
T4 and lambda, the genetics and biochemistry of eye colour, immunology,
and the basis of development in insects are just a short listing. The reason
for this synergism of approaches can be found in the elegant nature by
which a point mutation or small deletion precisely alters a phenotype,
which then is determined by a variety of biochemical technologies. Using
cross-feeding, genetic interaction, mUltiple mutant analysis, and precursor
accumulation data it was possible to develop, for instance, an integrated
idea of the development of bacteriophage assembly or tryptophan biosyn-
thesis. The same approach should be of value in the analysis of symbiotic
nodulation and nitrogen fixation.
The last two years have seen an escalation in the genetic analysis of the
legume. The most useful procedure was the direct selection for obvious
symbiotic phenotypes, such as non-nodulation or nitrate tolerance.
Induced mutagenesis has now been applied successfully to the soybean
(Glycine max), pea (Pisum sativum) and chickpea (Cicer arietinum). The fol-
lowing section will discuss recent advances in those three systems.
A. Pea and Chickpea Mutants
The essential features of these mutants were summarised by LaRue et al.

(1985) and Gresshoff et al. (1985). Davies et al. (1985) used gamma-ray irra-
diation of chickpea seeds, raised the first generation (termed the M!), and
then harvested the M2 seeds. These were planted out in large numbers and
screened for their inability to fix nitrogen. This was easily monitored, as
non-nodulated or non-fixing plants in the absence of exogenously supplied
nitrogen turned yellow, because of nitrogen starvation. Such yellow plants
were repotted and raised to maturity through the addition of nitrate fertil-
iser. If the plant failed to respond to nitrate application, it could be pre-
sumed that a general fitness gene, and not a symbiotic function, was
mutated. It was thus possible to eliminate chlorophyll deficiencies and
related deleterious mutations from the screen. Seeds from selected M3
plants were then tested for the presumed mutant phenotype. The chickpea
mutants isolated by Davies et al. (1985) included one non-nodulation and
four non-fixation mutants. All fell into separate complementation groups.
Two of the jix- mutants were temperature sensitive, as effective nodules
were formed at 24C but not at 29C. Wild-type plants formed effective
nodules at both temperatures but not at 34C. Mutant alleles rn4 and rns
determined the formation of ineffective nodules. One of these showed
nodule growth alterations, as nodules were small and white, while the other
(mutant PM796) had normal size nodules which were abnormally pale, i. e.
they were free of leghemoglobin. All mutant alleles in chickpea were
recessive. The non-nodulation mutant controlled by the rn! allele conferred
resistance to nodulation at all tested temperatures and with all tested
strains (Davies, 1985).
In P. sativum several research groups have taken advantage of the
elegant genetic system that this plant offers. The selection strategy in all
cases was the same as that described above. Kneen and LaRue (1984a, b)
described a non-nodulation mutant of the pea variety "Sparkle". Since
then they have isolated at least 30 further non-nodulation mutants using a
range of mutagens and have defined perhaps up to 10 separate comple-
mentation groups (LaRue et al., 1985; LaRue, pers. comm.). Perhaps a
word of warning should be included here regarding complementation
analysis. To achieve reliable data it is of importance to be certain that the
cross has indeed occurred. Thus the use of flanking markers or at least of
other donor pollen markers is essential. If such line differences have not
yet been bred into the relevant mutant material, then one requires at least a
several-fold repetition of the observation, before assigning markers to com-
plementation groups. Failure to complement may result from an incom-
plete cross so that the presumed hybrid seed actually stemmed from a self-
fertilisation. This would be easily confirmed in the subsequent generation.
This word of caution does not apply so much to research groups that are
familiar with plant breeding techniques, but to those that have only
recently entered the area. For example, all 15 mutants isolated by Messager
(1985; see later detailed description) fall into separate complementation
groups.
Feenstra and Jacobsen (1985) described the isolation of two non-nodu-

lation mutants in pea cultivar "Rondo". One mutant (K5) formed occa-
sional nodules on lateral roots, which were capable of efficient nitrogen
fixation. Root hair curling and infection thread formation were noted in
both the mutant and wild type, but occurred to a larger extent both in
degree and frequency over the root length of the wild-type plant. The
phenotype of mutant K5 was shown through grafting experiments to be
root-controlled. Another mutant isolated by Feenstra and Jacobsen was
mutant K24, which also was monogenic, but showed nodulation resistance
under all experimental conditions. Jacobsen et al. (1986) claimed that the
K24 mutation was both root- and shoot-controlled as illustrated by grafting
experiments with a nitrate-tolerant pea mutant. On the data supplied so
far, however, we feel that such interpretation may be premature and K24
might fit into the same category of root-controlled non-nodulation mutants
as previously described in pea (Kneen and LaRue, 1984 a, b).
Messager (1985) reported a large range of pea mutants with altered
symbiotic properties. A total of 7 nod- independent mutants were
described. All appeared to be monogenically inherited, with all of them
being recessive except for mutant FA.2l8, which was a dominant. Grafting
showed that non-nodulation was root-controlled. Eight mutants with inef-
fective symbioses were described. All were monogenic recessives. Ineffec-
tiveness like non-nodulation was root-controlled. One of the fix- mutants
(F261) showed conditional effectiveness, as bacterial strains from diverse
soils were isolated, which suppressed the symbiotic defect. Such condition-
ality was earlier pointed out in the naturally occurring soybean mutants,
and was described in pea with the "Afghanistan" line (Lie, 1971; Degen-
hardt et al., 1976) which failed to nodulate with most European R.
leguminosarum strains, but was nodulated effectively by isolates TOM and
HIM (Winarno and Lie, 1979). Messager (1985) furthermore described two
allelic mutants that nodulated in the presence of nitrate. They found nodu-
lation to be elevated by a factor of 3 to 6 compared to wild-type plants
grown under the same conditions. The term "hypernodulation" was used to
describe the phenomenon. The hypernodulation characteristic was asso-
ciated with fasciation in the flower of the pea plant. This might argue for
some pleiotropic effects, perhaps involving some growth regulator balance,
or a major deletion or genome rearrangement, which also affects this char-
acter. Jacobsen and Feenstra (1984) and Jacobsen (1984) described a
similar pea mutant which nodulated profusely in the presence of supplied
nitrate. No fasciation of the flower or stem was reported, although a
reduced ability to form a wild-type root system, even in the absence of
nodules, was noted. The mutant allele nod3 was recessive and was inherited
in a monogenic fashion. Grafting led Jacobsen to conclude that the nod3
phenotype was root-controlled. This conclusion may be entirely correct but
again, with the present set of supplied data, one has to query whether the
graft was made too late, and why the control plant was the non-nodulation
mutant K5 rather than the wild type. The hypernodulation mutant of Mes-
sager (1985) was shoot-controlled, which coincides with observations made
by Delves et al. (1986) and reported by Gresshoff et al. (1985). The nitrate
tolerant pea mutant of Jacobsen developed more than 250 nodules per
plant in the absence or presence of 15 mM nitrate, which in control plants
lowered the nodule number from about 59 to 17 per plant. The nod3 plants
likewise showed increased nitrogenase activity, as measured by acetylene
reduction, with about 5 micromoles ethylene being produced per hour per
plant with or without nitrate being present, but with nitrate lowering the
activity from 2 to 0.2 micromoles per hour per plant. Nodule fresh weight
per plant was nitrate inhibited (125 mg versus 26 mg), but not in the mutant
(589 mg versus 682 mg). The increase of nodule weight per plant on nitrate
underlined the ability of this mutant to develop a symbiosis in the presence
of the otherwise inhibitory levels of nitrate, and was caused by an increase
in total plant weight under nitrate. The latter point, and direct measure-
ments of nitrate reductase activity and nitrate content, demonstrated that
the nod3 mutation did not affect nitrate metabolism per se.
The pea system clearly provides several experimental advantages. The
genetic system is more defined than in any other legume, and there are
fewer linkage groups. Second, the plant is easy to grow under a variety of
conditions. Growth to maturity is fast. However, advantages relating
directly to the analysis of symbiotic nitrogen fixation are minimal, as the
symbiotic genetics of R. leguminosarum are not significantly ahead of
Bradyrhizobium japonicum. This was mainly achieved through the transfer
of "know-how" and gene-specific clones, which allowed rapid advance in
the slow-growing Bradyrhizobium species. Likewise one finds that the
knowledge of nodulin molecular biology, nodule biochemistry and the
developmental biology is nearly identical. Soybean has a large physio-
logical and biochemical data-base mainly because of the greater ease of
harvesting large amounts of nodule material. Determinate nodules, fur-
thermore, have the characteristic that the development of bacteroids or
nodule components occurs synchronously. Thus 13 day old soybean
nodules are relatively uniform in terms of their developmental stages,
whereas pea nodules contain a range of symbiotic stages with varying con-
tributions relative to the age of the nodule. The new researcher is advised
to look at those advantages in a plant system which are inherent biological
characteristics. Other disadvantages may only stem from a dearth of
research work with that class of organism. Often such short-falls can be
quickly made up through exchange of information and material.
B. Soybean Nodulation Mutants

The following are considerations in an induced mutagenesis programme.
The starting material should be homogeneous and homozygous. The plant
species must be self-fertile. Mutagenesis must achieve workable mutation
frequencies. There must be a convenient selection procedure.
The studies with soybean as described by Carroll et al. (1984, 1985 a, b;
Gresshoff et al., 1985 a, b) chose not to select for co-dominant or dominant
mutations in the M) generation, because the selection screen looked at
root-related phenotypes. The target tissue of induced mutagenesis is the

developing shoot meristem, which lies preformed in the dormant seed.
There is also a preformed root meristem, which upon germination will
develop into a chimeric root, provided that a stem cell in the meristem was
mutated. However, only mutations induced in the shoot meristem will be
passed on in the "germline" to the next generation. Thus any root charac-
teristic observed in the first generation of mutagenesis (M 1) would be ge-
netically dead and the M1 was screened only for the degree of fitness and
for chlorophyll deficient sectors; both parameters allowed the evaluation
of the mutagenic treatment. Three types of mutagen, gamma-rays, sodium
azide and ethyl methanesulphonate (EMS), were tried at varying doses.
Gamma-rays have great experimental advantages as seeds can be dry and
there are no carry-over effects to the experimenter. In our hands, both
gamma irradiation and chemical mutagenesis with sodium azide gave
increased mutation frequencies for chlorophyll deficiency above controls,
but these were only one-tenth of those observed with EMS. For optimal
EMS treatments, large lots of seeds were pregerminated in flywire bags in
well-aerated phosphate buffer at 28 C. For soybean it was necessary to
change the buffer as germination inhibitors were abundantly released
during pregermination. After 8 hours preincubation, EMS at 0.5 % (v/v)
was added. The EMS solution was drained from the seeds into an inacti-
vation vessel after 6 h and the seeds were further washed and rinsed prior
to wet planting in the soil. Direct planting seemed optimal as it was dif-
ficult to dry soybean seeds partially without severely affecting the fitness of
the seed. EMS in the inactivation vessel was inactivated by chemical
reaction. Detailed procedures of mutagenesis and safety procedures can be
obtained from us upon written request.
Optimal mutagenesis gave about 2.8 % chlorophyll deficiency in the M2
families. Carroll et al. (1985 a, b) collected individual M2 families, as this
allowed the later conclusion to be drawn that individual mutant isolates
were the product of separate mutagenic events. Additionally, one could
obtain a preliminary estimate of the mode of segregation. It was possible to
separate the mutant and wild-type segregants in the M 2, then to raise the
M3 and observe directly the stability or segregation of the mutant character.
It was thus feasible to evaluate the genetic nature of the individual mutant
alleles by the third generation after mutagenesis.
To select mutant phenotypes 12 seeds of each family were planted into
20 litre pots. The growth medium was river sand watered with nitrogen-
free growth medium plus Bradyrhizobium japonicum (strain
CB 1809 = USDA136) for selection of nod- and fix- plants, or medium con-
taining about 5 mM nitrate for the selection of nitrate-tolerant-symbiotic
(nts) mutants. An additional screen in flat trays was used to isolate soybean
mutants lacking the constitutive nitrate reductase activity (Carroll and
Gresshoff, 1986).
i) Constitutive Nitrate Reductase Deficiency
Nitrate was known to inhibit many phases of the symbiosis involving

legumes and Rhizobium (Carroll and Gresshoff, 1983). McNeil (1982)
found that different Rhizobium strains produced slight differences in the
symbiotic efficiency of soybean plants grown on nitrate. A similar
involvement of the Rhizobium was shown by Herridge et al. (1984), who
demonstrated that increased inoculant numbers were needed, if high
nitrate levels were present. However, since data by Gibson and Pagan
(1977) suggested that the plant contributed to the nitrate effect in a major
way, induced mutagenesis was used to isolate plant mutants which were
not affected by exogenous nitrate. Two strategies were utilised: one was to
isolate the phenotype directly (see later section VIII (iii)), and the other
was to attempt to isolate a blockage in nitrate uptake, which could be
detected as a nitrate reductase deficiency.
Similar mutants were isolated in Arabidopsis thaliana, although the
selection employed chlorate resistance screening (Braaksma, 1982).
However, soybean possesses at least two nitrate reductase activities
(Campbell, 1976); one inducible, NADH+ requiring, molybdenum-con-
taining enzyme similar to that found in other plants, and a constitutive
enzyme, which prefers NADPH+ and seems not to contain molybdenum.
Ryan et al. (1983) and Nelson et al. (1983) using chlorate selection isolated a
plant mutant which lacked constitutive nitrate reductase activity in the
soybean cultivar Williams. Chlorate acts as an analogue of nitrate and is
thought to act through its conversion by nitrate reductase to the toxic
chlorite ion. However, other interpretations of the chlorate effect also exist
(Cove, 1976). The soybean mutant LNR2 (defined by allele nrl) showed no
impairment of growth on nitrate and had identical symbiotic characteristics
when compared to the parental cultivar. It lacked the ability to release
nitrogenous compounds" which after catalytic conversion can be detected as
NO x in the form of nitrite. The NO x compound released during the in vivo
assay was identified as acetaldehyde oxime (Mulvaney and Hageman, 1984).
The correlation between constitutive nitrate reductase activity and NO x pro-
duction was 'tight, but conclusions were based on only the one allele.
Additional mutant isolates lacking this enzyme function are therefore
of importance to confirm or disprove the correlation. Until then, the phy-
siological function of the constitutive nitrate reductase of soybean will
remain unclear. Is it involved as a safety valve to high nitrate (or nitrite)
levels? Or is the measured activity an artifact of in vitro biochemistry and
does the enzyme have an alternative substrate in vivo? Spreit et al. (1985)
reported further complexities as the constitutive nitrate reductase activity
was separated into two molecular species. Thus the question arises whether
mutant LNR2 was a regulatory mutation affecting both peptides, or
whether there were multiple mutations (or large deletions), or whether the
two peptides were the product of the same gene. A mutant such as LNR2
should also be good starting material to select further mutations in the
inducible nitrate reductase activity of soybean.
The approach of Carroll and Gresshoff (1986) was to isolate a consti-

tutive nitrate reductase mutant by a direct screen rather than through
chlorate resistance. M2 plants were grown under stringent nitrate-free
conditions (to prevent the induction of the inducible nitrate reductase)
and discs from primary leaves were taken and assayed for nitrite pro-
duction using a colour assay in multi-titre plates. Putative mutants were
retested and grown to maturity to allow further characterisation of the
mutant character. Two mutants were isolated using this procedure (nr345
and nr328). Both had lowered enzymatic activity and their phenotype
was stable for at least 4 generations after isolation. Mutant nr328 had a
different phenotype than mutant LNR2, as it showed necrotic lesions on
leaves when grown on high (10 or 15 mM) nitrate (Whitmore-Smith,
1985). Symbiotically both mutants behaved similarly to wild-type Bragg
plants (Carroll, 1985).
ii) Non-Nodulation Mutants of Soybean
Three clear-cut non-nodulation mutants were obtained from mutant
screens of EMS-mutagenised Bragg seeds (Carroll et al., 1986). The iso-
lation procedure did not specifically focus on delayed nodulation mutants
although some of these may have been obtained. For example, in a mutant
screen for non-nodulation in cultivar Williams, six presumptive nod-
mutants segregated in M2 families, but upon retesting at maturity these
plants nodulated. Such mutants may define further nodulation genes which
are at present overlooked because the large screening programmes present
difficulties even for larger research groups. The analysis of the bacterial
nodulation genes has recently recognised the value of delayed nodulation
phenotypes. By this fashion extra nodulation genes were identified on the
Sym plasmid (Kondorosi et al., 1985).
Carroll et al. (1986) isolated mutants nod49, nod139 and nod772. All
grew well on nitrate media, indicating that the non-nodulation was not a
result of a general fitness mutation. Mutant nod49 had no pseudoinfection
or infection threads. Root hairs were not curled (W. D. Bauer and A.
Mathews, pers. comm.). Bradyrhizobium strain USDA76, which effectively
suppressed the rjl mutation in soybean, failed to do so with mutant nod49.
However, other Bradyrhizobium strains erratically caused nodules on
nod49 plants grown in Leonard jars (A. Mathews, pers. comm.). The
nodules were large (up to 113 mg per nodule compared to 11 mg per
nodule on wild-type plants), capable of efficient nitrogen fixation, and
were primarily positioned in the lower portions of the root system rather
than the crown region. Such increased nodule size was expected as the
plant compensated for the decreased nodule number (see Singleton and
Stockinger, 1983, for an illustration of compensation). This phenomenon,
already observed by Nutman (1953), implies that the developing meristem
of the nodule is controlled by systemic mechanisms, which recognise the
development of other nodule meristems in the plant - see also the effects
of the systemic signalling in split root systems as shown by Hinson (1975)
and Carroll and Gresshoff (1983). Whether the meristem is a source of
signal compounds or a sink is not known. Bacterial isolates from nodules

of nod49 yielded only the original inoculum strains suggesting that some
physiological, epigenetic phenomenon resulted in the successful invasion.
Whether this dealt with a plant (i. e. somatic) or bacterial effect was
unclear. In two separate field trials, mutant nod49 had remained non-nod-
ulated. In mutants nod49 and nod139 the root genotype controlled the
mutant phenotype (Delves et ai., 1986; Gresshoff et ai., 1985). Fur-
thermore, higher inoculation densities increased the nodule number found
on nod49 but not on plants of the parent cultivar Bragg grown in Leonard
jars. The mutant thus behaved in many ways like the rj1 allele in other
soybean cultivars. Allelism tests are in progress at the time of writing.
It is possible that the non-nodulation mutants described here are
altered in the production of root exudate substances. These substances
interact with the nodD gene product to activate a range of nodulation genes
in the bacterium (Halverson and Stacey, 1986). On the other hand it is fea-
sible that the nod- mutants are less sensitive to the relevant bacterial signal
which elicits the infection series. Or the plant mutant may have a hyper-
active defense response system which eliminates most bacterial interactions
at the infection stage. The ability of increased bacterial populations to
achieve some degree of nodulation on the otherwise non-nodulation
mutant nod49 suggests that this mutant fits into the latter two classes. In
part this is supported by data from plant co-culture (nod49 with Bragg),
which showed that the mutant phenotype was maintained despite close
proximity and general asepsis (Mathews, pers. comm.).
Halverson and Stacey (1984, 1985) found that a slow-to-nodulate
mutant of Bradyrhizobium japonicum strain USDAIlO could produce a
wild-type nodulation profile if preincubated with soybean seed lectin or
soybean root exudate. The tentative interpretation of this finding was that
the mutant was slow to interact with this exudate factor in vivo, and thus
some essential function could not be induced or suppressed. The lectin
bound to the bacterial surface may not necessarily help in the binding to
the plant cell surface (which may be a secondary property of bound lectin),
but rather makes the bacterium appear as if it is a non-pathogen (i. e. it is
camouflaged). Perhaps lectin bound to exopolysaccharides prevents the
recognition by the plant of LPS or other factors. Clearly the use of both
plant and bacterial mutants blocked in several stages of the infection
process will help in the biochemical solution to these questions. Preli-
minary analysis of the root exudate of seedlings of nod49 showed the
presence of identical amounts of soybean root lectin if compared to parent
cultivar Bragg or supernodulation mutant nts382 (see later section).
Carroll (1985) was also able to isolate some soybean mutants of cultivar
Bragg which were characterised by decreased nodule mass. These may fit
into the delayed nodulation category described above or may represent a
new phenotypic class in which the growth of the nodule is restricted
because of altered plant defense systems. This material has not been tested
further but could allow the biochemical analysis of nodule function and
nodule growth rather than nodule initiation.
iii) Nitrate-tolerant-symbiotic and Supernodulation Mutants of Soybean
Legumes utilise nitrate or atmospheric nitrogen gas as nitrogen source.

Usually they "prefer" soil nitrate, and symbiotic nitrogen fixation and nod-
ulation tend to be suppressed in the presence of nitrate. However, plants
tend to yield the same whatever the soil nitrate level provided sufficient
Rhizobium inoculant is present and environmental conditions favour good
symbiotic development. Thus the normal legume crop removes nitrate from
the soil, despite its natural ability to fix high amounts of nitrogen from the
atmosphere (Herridge et al., 1984). The mechanism of nitrate inhibition of
nodulation and nitrogen fixation has not been clarified. As a matter of fact,
one may have to consider multiple mechanisms in view of recent findings
in a number of plant species (Gibson and Harper, 1985). The isolation of
nitrate-tolerant legume mutants, as reported by Carroll et al. (1984, 1985 a,
b) and Gresshoff et al. (1985 a, b) for soybean, and Jacobsen and Feenstra
(1984), Jacobsen (1984) and Messager (1985) in pea, will aid with further
investigations.
Carroll et al. (1985 a, b) originally isolated 15 independent nts mutants
from EMS mutagenised Bragg families. Of these 12 were further charac-
terised. Table 1 shows a summary of their symbiotic properties in the
presence and absence of nitrate. Nitrate tolerant mutants have an increased
nodule number and mass under both culture conditions. Nitrogenase
activity in the mutants was significantly elevated in the presence of nitrate.
This was confirmed by total N determinations of different plant parts (D.
Day, pers. comm.). Mutants fell into different phenotypic classes as judged
by nodule number, nodule pattern, and plant fitness. Tentatively Gresshoff
et al. (1985b) have defined at least three complementation groups. Two
mutants (nts733 and nts 1116) were inherited as Mendelian co-dominant
alleles, while mutants nts382, nts501, nts2264, nts 1007 and nts 183 were
inherited as monogenic recessives. The mutants were described by the term
supernodulation and hypernodulation to indicate the degree of extra nodu-
lation compared to wild-type plants. Thus mutants nts382 and nts 1007
were clear supernodulation mutants, whereas mutant nts1116 and nts739
were more intermediate, hypernodulation types. Mutant nts382 grew on
nitrate and had nitrate reductase activity. This observation, together with
the increased nodule number in the absence of nitrate, led to the con-
clusion that the nts phenotype was due to supernodulation caused by a ces-
sation or diminishing of the autoregulation response.
The autoregulation response was described by Pierce and Bauer (1983)
on the tap root of soybean and by Kosslak and Bohlool (1984) on split
roots of soybean. It involves the regulation of nodule formation by other
nodulation events in a systemic fashion, so that ontogenetically younger
tissue close to the growing root tip recognises early nodulation stages in
more mature root tissue, leading to symbiotic arrest after the infection
stage (Calvert et al., 1984). Mutant nts382 and ntsl007 were tested in
plastic growth pouches and both showed an absence or lessening of the
autoregulation response. Mutant nts382 was further analysed in a quanti-
tative fashion, and Bauer (pers. comm.) demonstrated that the same
mechanism of infection was seen in the mutant as in the wild type. The
number of infection events (being the sum of all pseudoinfections and
infection thread structures) was increased, and the region of infectability
was maintained for a longer period of root development. Using growth
pouches, different soybean cultivars showed different degrees of autore-
gulation. In other words, the distribution of nodules was extended further
down the root in cultivar Bragg than in Williams. Preliminary results from
split root experiments involving Bragg and several other cultivars con-
firmed the single root observations (Bohlool and Gresshoff, unpublished
data).
Table 1. Summary of the symbiotic characteristics of soybean nts mutants
Soybean Nodule number Nodule dry weight Nitrogenase activity

genotype (mg) nmol C 2 H 4 min - I
- N0 3 +N0 3 - N0 3 +N03 -N03 +N0 3
Experiment 1
Bragg 26 (6) 19 (7) 34 (10) 5 (3) 71 (13) 1 (1)
nts382 576 (77) 1007 (154) 166 (9) 193 (20) 119 (35) 69 (11)
ntsl007 334 (292) 991 (231) 92 (49) 179 (35) 98 (29) 88 (20)
nts183 351 (71) 712 (188) 101 (10) 141 (37) 62 (8) 66 (7)
nts2264 478 (99) 907 (61) 124 (21) 188 (25) 99 (10) 55 (8)
ntslll6 101 (26) 74 (45) 66 (12) 30 (12) 85 (17) 23 (10)
nts733 457 (107) 797 (180) 115 (22) 172 (19) 88 (22) 54
Experiment 2
Bragg 15 (5) 15 (15) 47 (15) 4 (5) 112 (18) 3 (3)
nts501 253 (101) 235 (91) 115 (5) 55 (27) 173 (13) 20 (12)
nts2282 99 (67) 445 (82) 63 (20) 70 (29) 58 (21) 24 (13)
nts97 282 (88) 512(152) 87 (2) 92 (39) 72 (17) 28 (11)
nts739 94 (56) 175(84) 92 (21) 77 (24) 154 (52) 85 (41)
nts246 579 (157) 581 (162) 116 (43) 154 (45) 69 (21) 48 (8)
nts2062 123 (46) 299 (69) 120 (17) 144 (13) 169 (24) 108(53)
Plants were inoculated with Bradyrhizobium strain USOAII0 and harvested 26 days
after planting. Each entry is the mean of 3 to 6 plants (nodule number and dry
weight) or 2 plants (nitrogenase activity). SO are given in brackets. Carroll et ai.,
unpublished data.
Growth analysis of mutant nts382 in glasshouse conditions evidenced

that in the absence of Bradyrhizobium the shoot to root ratio of the nts
mutant versus the Bragg parent was altered during the first 50 days of
culture. For example, the shoot-root ratio of Bragg at 13 days was 5.5 com-
pared to 4.0 in nts382, but changed to 3.7 in Bragg versus 4.5 in nts382 after
50 days. Such uninoculated plants were grown on nitrate. Similar plants at

7 days showed that nts382 had more lateral roots (40 2 versus 29 2) and
similar tap root length (167 8 mm versus 167 8 mm) when compared to
Bragg. This was maintained at 14 days (845 mm versus 686 mm for
lateral root number and 246 15 mm versus 227 12 mm for tap root
length). (Day et a!., pers. comm.; Delves et ai., 1985.)
Not only were early nodulation initiation functions altered in mutant
nts382, but also late, i. e. nitrogen fixation related characteristics were dif-
ferent. For example, specific nitrogenase activity per nodule mass was
reduced to about 30 % of wild-type activity, but nitrogenase activity per
milligram bacteroid was equivalent (Delves et ai., 1986; Day, pers. comm.).
Mutant nts382 contained fewer bacteroids per unit nodule tissue. Mutant
ntsll16, however, which showed hypernodulation instead of supernodu-
lation, and which may be controlled by a different gene than nts382 and
nts 1007, showed intermediate amounts of bacteroid tissue to the parent.
Yet that mutant also had lowered specific nitrogenase activity. The causes
for these differences are totally unclear. The lowered bacteroid density was
confirmed by ultrastructural investigations of 30 day old nts382 nodules,
which were filled with only about one half of the bacteroids as wild type.
Additionally, the infected cells were considerably smaller showing less
Fig. 3. Root system of a supemodulating soybean plant grown under field condi-
tions. Mutant ntsl007 was inoculated with Bradyrhizobium USDAIIO and grown in
a moderately high nitrate soil. Irrigation was applied and the root system is shown
at maturity. Upto 3000 nodules can be obtained on such field-grown, well-inocu-
lated mutant plants. Control plants (not shown) have severe nodule senescence and
about 150 nodules per plant.
hypertrophism. Bacteroid number per peribacteroid vesicle was also

reduced. Mutant nts382 nodules resembled immature Bragg nodules or
Bragg nodules grown on nitrate. Occasional vacuoles were found in mutant
infected cells of young nodules, which were never seen in fully matured
wild-type cells (G. D. Price, pers. comm.).
Root systems of soil-grown mutant ntslO07 (tentatively allelic to mutant
nts382) showed strong supernodulation (Fig. 3), but the nodules had not
senesced at maturity compared to control Bragg plants.
Growth analysis of supernodulation mutants indicated that excessive
nodulation was associated with a reduction of shoot growth. Thus more
intermediate hypernodulators like mutant nts 1116 may be of greater
agronomic importance than mutant nts382. Day (pers. comm.) found that
lower bacterial inoculant densities also could result in decreased levels of
supernodulation and that in those cases the reduction of shoot weight was
not as pronounced.
In a small soil-grown growth trial Gresshoff et al. (1985b) found that
mutant nts 1007 grew nearly as well as the Bragg parent, although the root
system was heavily supernodulated (see Fig. 3 for a plant from that trial).
The major agronomic potential of the nitrate-tolerant plant mutants may
be their ability to fix nitrogen for a prolonged season in the presence of soil
nitrate. This nitrate could potentially be spared as the symbiotic depen-
Table 2. Summary of grafting data with supernodulation soybean mutants
Graft Nodule Nodule Tap root % root Nitrogenase

number mass (mg) length (cm) nodulated activity
Bragg 31 (II) 43 (10) 29 (5) upper33 1860 (360)

control
Bragg shoot/ 76 (26) 45 (21) 29 (5) upper33 1320 (780)
nts root
nts 278 (144) 151 (43) 18 (5) 95 480 (60)
control
nts shoot/ 213 (175) 109 (41) 23 (3) 95 1080 (300)
Bragg root
Data are means of 10 plants with the standard deviation in brackets. Nodule mass is dry
weight and nitrogenase activity is expressed as nmol ethylene produced per hour per gram
nodule dry weight. Nodule number and nodule mass values are expressed per gram dry
weight of plant.
dence of the plant is increased. This is of significance in rotational

cropping systems and possibly pastures, as nitrate needs to be available for
the companion or rotational species (see Tudge, 1984).
Reciprocal grafting of different mutant shoots onto wild-type root stock
showed that mutant nts382 (a supernodulation mutant) and mutant ntsll16
(a hypernodulation mutant, tentatively placed in a separate complemen-
tation group) controlled the root phenotype through a graft transmissible

shoot factor (Table 2, and Delves et al., 1986).
The combined data allowed the development of a general model to
explain the autoregulation response (Fig. 4). The model predicts that some
autoregulation mutants may be root-controlled. For this reason it is
essential that the apparently root-controlled nod3 mutant of pea is further
verified. The model also incorporates the results of split root studies of
Hinson (1975), Carroll and Gresshoff (1983) and Kosslak and Bohlool
(1984) in Fig. 5, and is expanded to involve a temporal sequence leading to
the establishment of nodulation and the autoregulatory sequence (Fig. 6).
Grafting eliminated flowers and cotyledons as the source tissue of
shoot-derived compound. Similarly approach grafts, in which a wild-type
and mutant shoot were grafted together onto the same root stock, demon-
strated that the mutant state could be associated with the absence of a
factor, as the root system always showed a wild-type nodulation response
(1. Olsson, pers. comm.). The observation also allows the expansion of this
technique, as one can now remove individual plant parts to see whether the
inhibitory signal from the wild-type plant can be removed. Also two nts
shoots could be grafted onto an nts root then various compounds could be
applied in bioassays to only one shoot.
nts mutant
X+W+V
!~ Me
: \., 1111
: a J..
I Elli
'--~~--1'~B
Fig. 4. A general model of autoregulation of nodulation in soybean. In wild-type
plants the shoot produces a signal (X) after being stimulated to do so by the root-
derived compound Q. Signal X acts as a inhibitor of symbiotic development at the
continued cell division stage. Compound Q in turn is a product of nodule meriste-
matic centers. In mutant nts382 and ntsl116 two different steps of the shoot
mechanism are disturbed. This results in a lesion of the regulatory loop as no X is
translocated to the root. Hence symbiotic development can continue towards
nodule formation. Basal levels of Q may also come from the other meristematic
centers of the root such as the root tip or cambial regions. Thus nts mutants, which
lack the ability to produce signal X may already be morphologically affected prior
to infection and cortical cell division. Nts mutant-wild type grafts mimick the
mutant situation, as no signal X is received by the root. Excessive nodulation may
increase levels of compound Q, so that eventually the signal level is sufficient to
produce some signal X, so that delayed inhibition of further nodulation occurs. RT
= root tip, N = nodule, B = bacterial invasion, + = positive activation, - =
inhibition. Me = meristematic centre, GJ = graft junction, Wand V are unknown
precursors of X.
In the natural situation, nodulation may be directly controlled by a phyto-

hormone balance between shoot and root system. Lawn and Bushby (1982)
studied the involvement of the root and shoot on nitrogen fixation in four
Vigna species which varied in their symbiotic potential. The shoot had a
pronounced effect on nodule fresh weight, while the root and the inoculant
were associated with the specific nodule activity. This was also reported by
Lawn and Brun (1974), who grafted between different soybean cultivars
and illustrated that ontogenetic change in the shoot might influence the
amount of photosynthate available for export to the root which influenced
the amount of nodule tissue formed. Root genotype had little effect on the
amount of nodule mass but did influence the ability of nodules to reduce
acetylene. Whether the interpretations regarding the involvement of photo-
synthate as a primary signal were correct is uncertain, especially in view of
the findings by Sheehy et al. (1980) and Williams et al. (1982), who showed
that there were no short-term increases in symbiotic nitrogen fixation of
soybean when whole plant photosynthesis was increased. Any increase in
the symbiotic tissue stemmed from an increased root system which mir-
rored the growth increases in the shoot. Additionally, such larger shoots
can be the sink for extra translocated nitrogen or they can be the source for
translocatable compounds other than sugars which influence the plant's
ability to regulate nodule development and growth.
The translocated compounds from shoots may include a range of
phytohormones. Very little is known about the involvement of phytohor-
mones in nodulation. Although Rhizobium synthesises cytokinins and
auxins (Phillips and Torrey, 1970), this may not be critical as many other
bacterial and fungal species produce such "common" metabolites as indole
acetic acid or adenine derivatives. Zeatin riboside is thought to be a major
form in which cytokinin moves from the root to the shoot (Summons et al.,
1981). Nodules were shown to be a sink for translocatable cytokinins
(Badenoch-Jones et al., 1983). Recently, Chen et al. (1985) reported that all
meristematic tissues in a plant were able to produce cytokinins. There are
suggestions that cytokinins move from the root tip to the developing
nodule. Thimann (1936) found that nodules contained high concentrations
of auxin. Indole acetic acid was recently unequivocally detected in culture
supernatants of Rhizobium. Mutants which lacked the ability to curl plant
root hairs still produced the same level of auxin as the parent strain
(Badenoch-Jones et al., 1982). Effective nodules of pea plants supplied at
their apical meristem with labelled auxin (indole acetic acid) accumulated
more radioactivity than ineffective nodules. Indole acetyl aspartic acid
appeared as the major metabolite (Badenoch-Jones et al., 1983). The
nodule tissue in all cases accumulated more auxin label than the corre-
sponding root tissue.
The nts mutants may possibly have an altered auxin biosynthetic
ability, so that the root system experiences increased cytokinin levels. This
may prevent hypertrophic extension of infected cells and may lead to
increased lateral root formation during early root growth, or increased nod-
ulation. While such speculations may be premature, it is worthwhile to
integrate these into models, as they aid experimental planning and hope-
fully stimulate discussion and feed-back. Of interest is the possible corre-
lation between auxin and nitrate metabolism. Tanner and Anderson (1963)
already suggested such links but focussed their thinking on the bacterial
side. It is possible that exogenous nitrate increases the effective auxin
levels in a particular tissue such as the root or the nodule. Such situations
may be the natural situation for all non-legumes which do not nodulate.
This may stimulate extensive root growth and may be involved in the sup-
pression of root exudate substances which may help to produce microbial
infestations of the root tissue. Likewise auxin may stimulate, either directly
or through a secondary molecule such as ethylene, phenolic biosynthesis
involved in the general anti-microbial mechanisms described previously.
A B
Fig. 5. The application of the model as shown in Fig. 4 to split root systems. At the
time of infection (A) on one side of the root, basal levels of compound Q from the
root tip stimulate basal levels of signal X release from the shoot. This level of X is
insufficient to prevent nodule initiation. Thus (B) nodule foci form, which develop
into nodules as shown in C. As sufficient meristematic centres develop in the cortex
of the one root half, higher amounts of compound Q are transported to the shoot,
resulting in the release of more signal X. This signal is transported to both root
halves (D), resulting in the suppression of new nodule development in the uninocu-
lated half and further nodulation in the already nodulated half (E). All symbols as
in Fig. 4.
x- w +- V shoot x-w-v
t:
'I
t:
"
"
: I basal '"I
I,
"I ~ :~
'~
Q '~
.........,
~..,......_ _-' root ~
URT RT
x-w-v
"': RT
Fig. 6. A temporal analysis of autoregulation events as predicted by the model out-
lined in Fig. 4. (1) Prior to infection, the shoot biosynthetic system leading to the
production of signal X is functioning at a basal level under basal stimulation of
compound Q, which is produced in the root. (2) Bacterial infection leads to initial
nodulation which comprises not only the formation of visible nodules but also of
meristematic centres in the root cortex. Not all of these regions will develop into
nodules. (3) These meristematic centres add to the level of compound Q, thereby
increasing the release of signal X from the shoot. (4) This results in a level of X in
the root which prevents any further symbiotic development and leads to symbiotic
arrest of the meristematic regions, so that a continued supply of compound Q is
produced; (5) thereby maintaining the inhibition of nodulation. Compound Q and
signal X may be more than one molecular species. Possible alternative sites for
mutation may be the inability to produce compound Q or the inability to release
signal X from the shoot. Some of these may be lethal and thus would be nearly
impossible to isolate without prior knowledge of their chemical requirements.
Signal X is also proposed to act in the general growth control of uninoculated
plants leading to altered shoot-root ratios and increased lateral root formation in its
absence as illustrated by the mutant nts382. All symbols are as in Fig. 4.
Whether nitrate achieves this by a stimulation of auxin synthesis in the

shoot, decreased breakdown in the root, an increased transport, or an
increased "susceptibility" of root tissues to the auxin signals is not known
(see Cross, 1985) for a review of auxin action).
Nitrate itself may be a "morphogen" and the nodulation response could
be blocked in its presence. Thus most non-legumes are "locked" in a nitrate
utilisation state and thus are "resistant" to nodulation by Rhizobium. For
example, nitrate may control cell wall biochemistry and thus the release of
specific oligosaccharides which may have morphogenetic effects, such as
those described by Tran et al. (1985). The described nts mutants may be
insensitive to the morphogenetic switch, which would explain the causes of
the nitrate tolerance.
In a related study Beach and Gresshoff (1986) found that roots of red
clover and siratro transformed by Agrobacterium rhizogenes were unable to
form nodules when inoculated with the Rhizobium. Yet mixed infections of
untransformed plants with Rhizobium and Agrobacterium at equal titres
gave no inhibition. Perhaps the phytohormone changes associated with the
transformed state prevented symbiotic development.
These speculations show how a few mutants can open up a range of dif-
ferent experimental approaches. More mutants will be isolated in a wider
range of legumes and the existing material will be described in a large
variety of ways. For example, Bisseling (pers. comm.) looked at in vitro
translation products of total RNA isolated from 5 day old nts382 and
Bragg roots by 2 dimensional electrophoresis, and found that the mutant
root lacked one peptide spot of molecular weight of about 37000. Why
such a difference existed, when the regulation of supernodulation in this
mutant was shoot-controlled, seems unclear, but may suggest that some
secondary molecular changes were already in place at this early stage of
root development prior to nodulation.
IX. Conclusions
The nodulation of legumes is characterised by a wide variety of

mechanisms showing immense ability for adaptation. Within this range of
phenotypes there are, however, general principles. The analogy to leaves
and photosynthesis comes to mind. Despite a seemingly unlimited range of
form and environmental optimisation, most photosynthetic and gas
exchange processes fall into two, possibly three, different categories. The
same balance between diversity and stability is found in nodulation
biology. The existence of mutants altered in the nodulation and nitrogen
fixation capacities indicated that similar phenotypes, tissue control and
genes of similar genetic function were detectable in a variety of legumes.
This commonality spanned even the gap between the determinate and inde-
terminate nodule type. Fundamental functions necessary for nodulation
control are similar between all legumes (and perhaps even non-legumes).
Micro-evolution may have inserted numerous fine adjustments giving the
range of symbioses seen today. Leghemoglobin amino-acid sequences can

diverge by as much as 50 %, yet all molecules are still capable of fulfilling
their symbiotic function of oxygen transport. Similar changes could have
occurred in other genes. The commonality of the host defense systems is
amazing, as is the similarity of the genes and mechanisms involved in other
plant-bacterial interactions such as those described for Agrobacterium or
pathogenic Pseudomonas species.
The biochemical analysis of the plant responses in symbiotic nodu-
lation and nitrogen fixation has just started. Hopefully, with the new set of
tools available to scientists, continued analysis will not be hampered, since,
although seemingly important in crop productivity, not all discoveries will
result in commercial advances. Despite the large data-base in photosyn-
thesis, there have been no commercial applications of this knowledge. The
study of plant development and plant biochemistry, however, goes beyond
the commercial. Hopefully the coupling of plant breeding, plant cell
culture, molecular genetics and classical biochemistry can help in the eluci-
dation of the complex developmental processes in plants, which form the
basis of all life on this planet.
Acknowledgements
The authors thank Dr. Bernard J. Carroll, Dr. David Day and Dr. Dean
Price as well as our graduate students for extensive discussions and the
provision of unpublished material. Special thanks also to the Genetics
Department and Professor Alfred Piihler (University of Bielefeld), where
the major part of the writing was completed. Dr. Wolfgang D. Bauer, Dr.
Ben Bohlool and Dr. Ton Bisseling are thanked for providing additional
information about the nts mutants. The Alexander von Humboldt Found-
ation and Agrigenetics Corporation are thanked for direct and indirect
support.
x. References
Albersheim, P., Anderson-Prouty, A J., 1975: Carbohydrates, proteins, cell surface
and the biochemistry of pathogenesis. Annu. Rev. Plant Physiol. 26, 31-52.
Appleby, C. A, 1984: Leghemoglobin and Rhizobium respiration. Annu. Rev. Plant
Physiol. 35, 443-478.
Appleby, C. A, 1985: Plant hemoglobin: properties, function and genetic origin.
In: Ludden, P. W., Burris, J. E. (eds.), Nitrogen Fixation and CO 2 Metabolism,
pp. 41-51. Amsterdam: Elsevier.
Appleby, C. A, Tjepkema, J. D., Trinick, M. J., 1983: Hemoglobin in the non-legu-
minous plant Parasponia. Possible genetic origins and function in nitrogen
fixation. Science 220, 951-953.
Ausubel, F., Buikema, W., Earl, c., Klingensmith, J., Nixon, B. T., Szeto, W., 1985:
Organisation and regulation of Rhizobium meliloti and Parasponia Bradyrhi-
zobium nitrogen fixation genes. In: Evans, H. J., Bottomley, P. J., Newton, W.
E. (eds.), Nitrogen Fixation Research Progress, pp. 165-171. Dordrecht:
Nijhoff Publ.
Badenoch-Jones, J., Summons, R., Djordjevic, M. A., Shine, J., Letham, D. S.,
Rolfe, B. G., 1982: Mass-spectrometric quantification of indole-3-acetic acid in
Rhizobium culture supernatants: Relation to root hair curling and nodule
initiation. AppI. Environ. MicrobioI. 44, 27S-280.
Badenoch-Jones, J., Rolfe, B. G., Letham, D. S., 1983: Phytohormones, Rhizobium
mutants and nodulation in legumes. III.: Auxin metabolism in effective and
ineffective pea root nodules. Plant PhysioI. 73, 347 -3S2.
Barton, K. A., Brill, W. J., 1983: Prospects in plant genetic engineering. Science 219,
671-676.
Bassam, B., Gresshoff, P. M., 1986: Use of neomycin for preferential selection
against Rhizobium trifolii in symbiosis with white clover (Trifolium repens). Aust.
J. BioI. Sci. 37, in press.
Bauer, W. D., 1981: Infection oflegumes by rhizobia. Annu. Rev. Plant PhysioI. 32,
407-449.
Beach, K., Gresshoff, P. M., 1986: Culture and nodulation properties of legume
roots transformed by Agrobacterium rhizogenes. Plant Science (submitted).
Bedmar, E. J., Phillips, D. A., 1983: Pisum sativum cultivar effects on hydrogen
metabolism of Rhizobium. Can. J. Bot. 62, 1682-1686.
Bedmar, E. J., Phillips, D. A, 1984: A transmissible plant shoot factor promotes
uptake hydrogenase activity in Rhizobium symbionts. Plant PhysioI. 75,
629-633.
Bedmar, E. J., Edie, S. A, Phillips, D. A, 1983: Host plant cuItivar effects on
hydrogen evolution by Rhizobium leguminosarum. Plant PhysioI. 72,
lOll-lOIS.
Bergersen, F. J., Goodchild, D. J., 1973: Aeration pathways in soybean root
nodules. Aust. J. BioI. Sci. 26, 729-740.
Bergersen, F. J., Turner, G. L., 1967: Nitrogen fixation by the bacteroid fraction of
breis of soybean root nodules. Biochim. Biophys. Acta 141, S07-S1S.
Bergersen, F. J., Turner, G. L., 1978: Activity of nitrogenase and glutamine syn-
thetase in relation to availability of oxygen in continuous cultures of a strain of
cowpea Rhizobium species supplied with excess ammonium. Biochim. Biophys.
Acta 538, 406-41S.
Bergmann, H., Preddie, E., Verma, D. P. S., 1983: Nodulin 3S: A subunit of specific
uricase (uricase II) induced and localized in uninfected cells of nodules. EMBO
J. 2, 2333-2339.
Billing, E., 1982: Entry and establishment of pathogenic bacteria in plant tissues.
In: Rhodes-Roberts, M. E., Skinner, F. A (eds.), Bacteria and Plants, pp.
SI- 70. London: Academic Press.
Bisseling, T., Been, C., Klugkist, J., van Kammen, A., Nadler, K., 1983: Nodule spe-
cific host proteins in effective and ineffective root nodules of Pisum sativum.
EMBO J. 2, 961-966.
Bisseling, T., Govers, F., Gloudemans, T., Moerman, M., van Kammen, A., 1985:
Expression of pea nodulins in effective and ineffective symbiosis. In: Analysis
of the Plant Genes Involved in the Legume-Rhizobium Symbiosis, pp. 104-111.
Paris: OECD PubI.
Bowen, G. D., Kennedy, M. M., 1961: Heritable variation in nodulation of Cen-
trosema pubescens Beuth. Qld. J. agri. Sci. 18, 161-170.
Braaksma, F., 1982: Genetic control of nitrate reduction in Arabidopsis thaliana.
PhD dissertation Univ. of Groningen, Haren, The Netherlands.
Brill, W. J., 1980: Biochemical genetics of nitrogen fixation. MicrobioI. Rev. 44,
449-467.
Caldwell, B. E., 1966: Inheritance of a strain specific ineffective nodulation in soy-

beans. Crop Sci. 6, 427 -428.
Calvert, H. E., Pence, M., Pierce, M., Malik, N. S. A, Bauer, W. D., 1984: Ana-
tomical analysis of the development and distribution of Rhizobium infections in
soybean roots. Can. 1. Bot. 62, 2375-2384.
Campbell, W. H., 1976: Separation of soybean leaf nitrate reductases by affinity
chromatography. Plant Sci. Lett. 7, 239-247.
Carroll, B. 1., 1985: The plant contribution to the soybean-Rhizobium symbiosis.
PhD dissertation, Australian National University, Canberra, Australia.
Carroll, B. 1., Gresshoff, P. M., 1983: Nitrate inhibition of nodulation and nitrogen
fixation in white clover. Z. Pflanzenphysiol. 110, 77 -88.
Carroll, B. 1., Gresshoff, P. M., 1986: Isolation after induced mutagenesis and
initial characterisation of constitutive nitrate reductase mutants NR328 and
NR345 of soybean. Plant Physiol., in press.
Carroll, B. 1., McNeil, D. L., Gresshoff, P. M., 1984: Breeding soybeans for
increased nodulation in the presence of external nitrate. In: Ghai, B. S. (ed.),
Symbiotic Nitrogen Fixation, Vol. I, pp 43-50. Ludhiana, India: USG Publ.
Company.
Carroll, B. 1., McNeil, D. L., Gresshoff, P. M., 1985a: Isolation and properties of
soybean (Glycine max) mutants that nodulate in the presence of high nitrate
concentrations. Proc. Nat. Acad. Sci. U.S.A 82,4162-4166.
Carroll, B. 1., McNeil, D. L., Gresshoff, P. M., 1985b: A supernodulation and
nitrate tolerant symbiotic (nts) soybean mutant. Plant Physiol. 78, 34-40.
Carroll, B. 1., McNeil, D. L. Gresshoff, P. M., 1986: Isolation of non-nodulation
mutants of soybean by induced mutagenesis. Plant Science (submitted).
Chen, C. M., Ertl, 1. R., Leisner, S. M., Chang, C. c., 1985: Localization of cyto-
kinin biosynthetic sites in pea plants and carrot roots. Plant Physiol. 78,
510-513.
Clark, F. E, 1957: Nodulation responses in two near isogenic lines of the soybean.
Can. 1. Microbiol. 3, 113-123.
Collins, 1., 1983: Anatomical investigations of nodule initiation in white clover.
Honours Degree Dissertation, Botany Department, Australian National Uni-
versity, Canberra.
Cove, D. 1., 1976: Chlorate toxicity in Aspergillus nidulans: studies of mutants
altered in nitrate assimilation. Mol. Gen. Genet. 146, 147-159.
Cross, 1. W., 1985: Auxin action: the search for the receptor. Plant, Cell Environ. 8,
351-359.
Cullimore, 1. V., Lara, M., Lea, P. 1., Miflin, B. 1., 1983: Purification and properties
of two forms of glutamine synthase from the plant fraction of Phaseolus root
nodules. Planta 157, 245-253.
Daly, 1. M., 1984: The role of recognition in plant disease. Annu. Rev. Phytopathol.
22,273-307.
Davies, T. M., 1985: Host genes affecting nodule formation and function in
chickpea (Cicer arietinum). In: Evans, H. 1., Bottomley, P. 1., Newton, W. E.
(eds.), Nitrogen Fixation Research Progress, p40. Dordrecht: Nijhoff Publ.
Davies, T. M., Foster, K. W., Phillips, D. A, 1985: Non-nodulation mutants of
chickpea. Crop Sci. 25, 345-348.
Day, P. R., 1974: Genetics of host-parasite interaction. San Francisco: W. H.
Freeman.
Dazzo, F. B., Gardiol, A E., 1984: Host specificity in the Rhizobium-legume inter-
action. In: Verma, D. P. S., Hohn, Th. (eds.), Genes Involved in the Microbe-
Plant Interaction. Plant Gene Research. Vol. 1, pp. 3-31. Wien - New York:
Springer
Degenhardt, T. L., LaRue, T. A, Paul, E. A, 1976: Investigation of a non-nodu-
lating cultivar of Pisum sativum. Can. J. Bot. 54, 1633-1636.
Delannay, X., Rodgers, D. M., Palmer, R. G., 1983: Relative genetic contributions
among ancestral lines to North American soybean cultivars. Crop Sci. 23,
944-949.
Delves, A C., Day, D. A, Price, G. D., Carroll, B. J., Gresshoff, P. M., 1985: Regu-
lation of nodulation and nitrogen fixation in nitrate tolerant, supernodulating
soybeans. In: Evans, H. J., Bottomley, P. J., Newton, W. E. (eds.), Nitrogen
Fixation Research Progress, p41. Dordrecht: Nijhoff Publ.
Delves, A. c., Mathews, A, Day, D. A, Carter, A C., Carroll, B. J., Gresshoff, P.
M., 1986: Regulation of the Rhizobium-legume symbiosis by root and shoot
factors (submitted).
Devine, T. E., 1985: Nodulation of soybean plant introduction lines with fast-
growing rhizobial strain USDA 205. Crop Sci. 25, 354-356.
Devine, T. E., Weber, D. F., 1977: Genetic specificity of nodulation. Euphytica 26,
527-535.
Devine, T. E., Palmer, R G., Buzzell, R. I., 1983: Analysis of genetic linkage in the
soybean. J. Heredity 74, 457 -460.
Ditta, G., Corbin, D., Leong, S., Barran, L., Helinski, D. R, 1983: Symbiotic
nitrogen fixation genes of Rhizobium meliloti. In: PUhler, A (ed.), The
molecular genetics of the bacteria-plant interaction, pp. 88-97. New
York - Berlin - Heidelberg: Springer.
Djordjevic, M. A, Zurkowski, W., Shine, J., Rolfe, B. G., 1983: Sym plasmid
transfer to various symbiotic mutants of Rhizobium trifolii, Rhizobium legumin-
osarum and Rhizobium meliloti. J. Bacteriol. 156, 1035-1045.
Dommergues, Y. R., 1978: Impact on soil management and growth. In: Dom-
mergues, Y. R, Krupa, S. V. (eds.), Interactions Between Non-pathogenic Soil
Microorganisms and Plants, pp. 443-458. Amsterdam: Elsevier.
Duke, S. H., Henson, C. A., 1985: Legume nodule carbon utilization in the syn-
thesis of organic acids for the production of transport ami des and amino acids:
In: Ludden, P. W., Burris, J. E. (eds.), Nitrogen Fixation and CO 2 Metabolism,
pp. 293-302.Amsterdam: Elsevier.
Elkan, G. H., 1961: A nodulation inhibiting root excretion from a non-nodulating
soybean strain. Can. J. Microbiol. 7, 851-856.
Elmerich, C., 1984: Molecular biology and ecology of diazotrophs associating with
non-leguminous plants. Biotechnology 2, 967 -978.
Eskew, D. L., Schrader, L. E., 1977: Effect of rjl rjl (non-nodulating) soybeans on
nodulation of near isogenic RjlRjl plants in nutrient culture. Can. J. Microbiol.
23, 988-993.
Evans, L. S., Van't Hof, J., 1973: Cell arrest in G2 in root meristems: a control
factor from the cotyledons. Exp. Cell Res. 82, 471-473.
Feenstra, W. J., Jacobsen, E., 1985: Pea mutants with an altered response to Rhi-
zobium leguminosarum. In: Analysis of Plant Genes Involved in the Legume-
Rhizobium Symbiosis, pp. 50-51. Paris: OECD Publ.
Fischer, H. M., Hennecke, H., 1984: Linkage map of Rhizobium japonicum nifH and
nijDK operons encoding the polypeptides of the nitrogenase enzyme complex.
Mol. Gen. Genet. 196,537-540.
Flor, H. H., 1955: Host-parasite interaction in flax rust - its genetic and other
implications. Phytopathol. 45, 680-685.
Flor, H. H., 1971: Current status of the gene-for-gene concept. Annu. Rev. Phyto-
pathol. 9, 275-296.
Fraser, R. S. S., 1982: Are "pathogenesis-related" proteins involved in acquired sys-
temic resistance of tobacco plants to tobacco mosaic virus? J. Gen. Virol. 58,
305-313.
Fuller, F., Verma, D. P. S., 1984: Appearance and accumulation of nodulin mRNAs
and their relationship to the effectiveness of the root nodule. Plant Mol. BioI. 3,
21-28.
Fuller, F., Kiinster, P. N., Nguyen, T., Verma, D. P. S., 1983: Soybean nodulin
genes: analysis of cDNA clones reveals several major tissue specific sequences
in nitrogen fixing root nodules. Proc. Nat. Acad. Sci. U.S.A. 80, 2594-2598.
Gadal, P., 1983: Phosphoenolpyruvate carboxylase and nitrogen fixation. Physiol.
veg. 21, 1069-1074.
Gelin, 0., Blixt, S., 1964: Root nodulation in peas. Agric. Hort. Genet. 22,
149-159.
Gibson, A. H., Harper, J. E., 1985: Nitrate effects on nodulation of soybean by Rhi-
zobiumjaponicum. Crop Sci. 25, 497-501.
Gibson, A. H., Pagan, J., 1977: Nitrate effects on the nodulation of legumes inocu-
lated with nitrate reductase deficient mutants of Rhizobium. Planta 134, 17-32.
Gorbet, D. W., Burton, J. C., 1979: A non-nodulating peanut. Crop Sci. 19,
727-728.
Govers, F., Gloudemans, T., Moerman, M., van Kammen, A., Bisseling, T., 1985:
Expression of plant genes during the development of pea root nodules. EMBO
J. 4, 861-867.
Gresshoff, P. M., Mohapatra, S. S., 1981: Legume cell and tissue culture. In: Rao,
A. N. (ed.), Tissue Culture of Economically Important Crop Plants, pp. 11-24.
Singapore: Univ. Singapore Press.
Gresshoff, P. M. and 14 other authors, 1984: The Parasponia-Rhizobium nitrogen
fixing symbiosis: Genetics, biochemistry, and molecular biology of a plant and
bacterium. In: Swaminathan, M. (ed.), Genetics: New Frontiers, pp. 217-226.
New Delhi: Oxford and IBH Publ. Compo
Gresshoff, P. M., McNeil, D. L., Carroll, B. J., 1985 a: Symbiotic mutants of
soybean capable of supernodulation in the absence and presence of nitrate. In:
Szalay, A. A., Legocki, R. P. (eds.), Advances in the Molecular Genetics of the
Bacteria-plant Interaction, pp. 49-51. Ithaca, N. Y.: Cornell Univ. Publ.
Gresshoff, P. M., Day, D. A., Delves, A. C., Mathews, A., Olsson, J. A., Price, G.
D., Schuller, K. A., Carroll, B. J., 1985b: Plant host genetics of symbiotic
nitrogen fixation and nodulation in pea and soybean. In: Evans, H. J., Bot-
tam ley, P. J., Newton, W. E. (eds.), Nitrogen Fixation Research Progress,
pp. 19-25. Dordrecht: Nijhoff Publ.
Guerinot, M. L., Chelm, B., 1985: Bacterial cAMP and heme in the Rhizobium-
legume symbiosis. In: Evans, H. J., Bottomley, P. J., Newton, W. E., (eds.),
Nitrogen Fixation Research Progress, p. 220. Dordrecht: Nijhoff Publ.
Halverson, L. J., Stacey, G., 1984: Host recognition in the Rhizobium-soybean sym-
biosis: Detection of a protein factor in soybean root exudate which is involved
in the nodulation process. Plant Physiol. 74, 84-89.
Halverson, L. J., Stacey, G., 1985: Host recognition in the Rhizobium-soybean sym-
biosis: Evidence for the involvement of lectin in nodulation. Plant Physiol. 77,
621-625.
Halverson, L. J., Stacey, G., 1986: Plant-microbial interaction, Microbiol. Rev., in
press.
Hanks, J. F., Schubert, K., Tolber, N. E., 1983: Isolation and characterization of
infected and uninfected cells from soybean nodules. Role of uninfected cells in
ureide synthesis. Plant Physiol. 71, 869-873.
Harper, J. E., Gibson, A H., 1984: Differential nodulation tolerance to nitrate
among legume species. Crop Sci. 24, 797-801.
Heath, M. C., 1980: Reactions ofnonsuscepts to fungal pathogens. Annu. Rev. Phy-
topathol. 18,211-236.
Heath, M. c., 1981: A generalized concept of host-parasite specificity. Phytopathol.
71, 1121-1123.
Herridge, D. F., 1982: Use of the ureide technique to describe the nitrogen
economy of field-grown soybean. Plant Physiol. 70, 7 -11.
Herridge, D. F., Betts, J., 1985: Nitrate tolerance in soybean. Variation between
genotypes. In: Evans, H. J., Bottomley, P. J., Newton, W. E. (eds.), Nitrogen
Fixation Research Progress, p. 32. Dordrecht: Nijhoff Publ.
Herridge, D. F., Rough1ey, R. J., Brockwell, J., 1984: Effect of rhizobia and soil
nitrate on the establishment and functioning of the soybean symbiosis in the
field. Aust. J. Agric. Res. 35, 149-161.
Hinson, K., 1975: Nodulation response from nitrate applied to soybean half-root
systems. Agron. J. 67, 799-804.
Hodgson, A C. M., Stacey, G., 1986: Potential for Rhizobium improvement. Crit.
Rev. in Biotechnol., in press.
Holl, F. B., 1975: Host plant control of the inheritance of dinitrogen fixation in the
Pisum sativum-Rhizobium symbiosis. Euphytica 24, 767 - 770.
Holl, F. B., 1983: Plant genetics: manipulation of the host. Can. J. Microbiol. 29,
945-953.
Holl, F. B., LaRue, T. A, 1976: Host genetics and nitrogen fixation. In: Hill, L. D.
(ed.), Proc. World Soybean Conf. Urbana, Ill., pp. 156-163. Danville: The
Interstate Printers and Publishers.
Holl, F. B., Milliron, M. L., Delafield, S. J., 1983: Quantitative variation in root
nodule 1eghemoglobins: Interspecific variation in component problems. Plant
Sci. Lett. 32, 321-326.
Jacobsen, E., 1984: Modification of symbiotic interaction of pea (Pisum sativum) and
Rhizobium /eguminosarum by induced mutation. Plant and Soil 82, 427 -438.
Jacobsen, E., Feenstra, W. J., 1984: A new pea mutant with efficient nodulation in
the presence of nitrate. Plant Sci. Lett. 33, 337 -344.
Jacobsen, E., Postma, J. G., Nijdam, H., 1985: Genetical and grafting experiments
with pea mutants in studies of symbiosis. In: Evans, H. J., Bottomley, P. J.,
Newton, W. E. (eds.), Nitrogen Fixation Research Progress, p. 43. Dordrecht:
Nijhoff Publ.
Johansen, E., Finan, T. M., Gefter, M. L., Signer, E. R., 1984: Monoclonal anti-
bodies to Rhizobium meliloti and surface mutants insensitive to them. J. Bac-
teriol. 160,454-457.
Jones, G. D., 1962: Variation in nodule characters in S. 100 Nomark white clover. J.
Sci. Food. Agric. 13, 598-603.
Jordan, D. C., 1982: Transfer of Rhizobiumjaponicum Buchanan 1980 to Bradyrhi-
zobium japonicum gen. nov., a genus of slow-growing, root nodule producing
bacteria from leguminous plants. Int. J. Syst. Bacteriol. 32, 136-139.
Kahn, M., Krans, J., Sommerville, J. E. 1985. A model of nutrient exchange in the
Rhizobium-legume symbiosis. 1985: In: Evans, H. J., Bottomley, P. J., Newton,
W. E. (eds.), Nitrogen Fixation Research Progress, pp. 193-199. Dordrecht:
Nijhoff Publ.
Kaninakis, P., Verma, D. P. S., 1985: Nodulin-24 gene of soybean encodes for a
peptide of the peribacteroid membrane and was generated by tandem dupli-
cation of a sequence resembling an insertion element. Proc. Nat. Acad. Sci.
U.S.A 82,4157-4161.
Keen, N. T., Kennedy, B. W., 1974: Hydroxyphaseolin and related isoflavonoids in
the hypersensitive resistant reaction of soybean against Pseudomonas glycinea.
Physiol. Plant Pathol. 4, 173-185.
Keen, N. T., Williams, P. H., 1971: Chemical and biological properties of a lipomu-
copolysaccharide from Pseudomonas lachrymans. Physiol. Plant Pathol. 1,
247-264.
Keithley, J. H., Nadler, K. D., 1983: Protoporphyrin formation in Rhizobium
japonicum. J. Bacteriol. 154, 838-845.
Klement, Z., 1982: Hypersensitivity. In: Mount, M. S., Lacey, G. H. (eds.), Phyto-
pathogenic Prokaryotes. Vol. 2., pp. 150-177. New York: Academic Press.
Kneen, B. E., LaRue, T. A, 1984a: Nodulation resistant mutant of Pisum sativum
(L). J. Hered. 75, 238-240.
Kneen, B. E., LaRue, T. A., 1984b: Peas (Pisum sativum L.) with strain specificity
for Rhizobium leguminosarum. Heredity 52, 383-389.
Kondorosi, A, and 11 other authors, 1985: Identification and organisation of Rhi-
zobium meliloti genes relevant to the initiation and development of nodules. In:
Evans, H. J., Bottomley, P. J., Newton, W. E. (eds.), Nitrogen Fixation Research
Progress, pp. 73-78. Dordrecht: Nijhoff Publ.
Kondorosi, E., Baufalvi, Z., Kondorosi, A, 1984: Physical and genetical analysis of
a symbiotic region of Rhizobium meliloti. Identification of nodulation genes.
Mol. Gen. Genet. 193,445-452.
Kortt, A A, Burns, J. E., Trinick, M. J., Appleby, C. A, 1985: The amino-acid
sequence of hemoglobin I from Parasponia anderson ii, a non-leguminous plant.
FEBS Lett. 180, 55-60.
Kosslak, R. M., Bohlool, B. B., 1984: Suppression of nodule development of one
side of a split root system of soybeans caused by prior inoculation of the other
side. Plant Physiol. 75, 125-130.
LaFavre, J. S., Eaglesham, A R. J., 1984: Increased nodulation of "non-nodu-
lating" (rjlrjl) soybeans by high dose inoculation. Plant and Soil 80,297-300.
Lalonde, M., 1979: Immunological and ultrastructural demonstration of nodulation
of the European Alnus glutinosa (L) Gaerthn. host plants by actinomycetal iso-
lates from the North American Compteria peregrina (L.) Coult root nodule. Bot.
Gaz. (suppl.) 140, 535-543.
Lang-Unnasch, N., Ausubel, F. M., 1985: Nodule-specific polypeptides from
effective alfalfa root nodules and from ineffective nodules lacking nitrogenase.
LaRue, T. A., Kneen, B. E., Gartside, E., 1985: Plant mutants defective in symbiotic
nitrogen fixation. In: Analysis of the Plant Genes Involved in the Legume-Rhi-
zobium Symbiosis, pp. 39-48. Paris: OECD Publ.
Lawn, R. J., Brun, W. A., 1974: Symbiotic nitrogen fixation in soybean III: Effect
of supplement nitrogen and intervarietal grafting. Crop Sci. 14, 22-25.
Lawn, R. J., Bushby, H. V. A, 1982: Effect of root, shoot and Rhizobium strain on
nitrogen fixation in four aseatic Vigna species. New Phytol. 92,425-434.
Lee, J. S., Brown, G. G., Verma, D. P. S., 1983: Chromosomal arrangement of leg-
hemoglobin genes in soybean. Nuc. Acids Res. 16,5541-5553.
Legocki, R. P., Verma, D. P. S., 1979: A nodule specific plant protein (nodulin 35)
from soybean. Science 205, 190-193.
Legocki, R P., Eaglesham, A. R 1., Szalay, A. A., 1983: Stem nodulation in Aes-
chynomene: A model system for bacteria-plant interaction In: PUhler, A. (ed.),
The molecular genetics of the bacteria-plant interaction, pp. 210-219. New
York - Berlin - Heidelberg: Springer.
Leong, S. A., Ditta, G. S., Helinski, D. R, 1982: Heme biosynthesis in Rhizobium:
identification of a cloned gene coding for gamma-aminolevulinic acid synthase
in Rhizobium meliloti. 1. BioI. Chern. 257, 8724-8730.
Lie, T. A., 1971: Symbiotic nitrogen fixation under stress conditions. Plant and Soil,
special volume, 117-127.
Masterson, R V., Prakash, R K., Atherly A. G., 1985: Conservation of symbiotic
nitrogen fixation gene sequences in Rhizobium japonicum and Bradyrhizobium
japonicum. 1. Bacteriol. 163, 21-26.
McNeil, D. L., 1982: Variation in the ability of Rhizobium strains to nodulate soy-
beans and maintain fixation in the presence of nitrate. Appl. Environ. Mi-
crobiol. 44, 647-652.
McNeil, D. L., LaRue, T. A., 1984: Effect of nitrate source on ureides in soybean.
Plant PhysioL 74, 227 -:-232.
McNeil, D. L., Carroll, B. 1., Gresshoff, P. M., 1984: The nitrogen fixation capacity
of bacteroids extracted from soybean nodules inhibited by nitrate, ammonia
and dark treatments. In: Ghai, B. S. (ed.), Symbiotic Nitrogen Fixation, Vol. I,
pp. 79-88. Ludhiana, India: USG Publ.
Messager, A., 1985: Mutants for nodulation and nitrogen fixation. In: Analysis of
the Plant Genes Involved in the Legume-Rhizobium Symbiosis, pp. 52-60.
Paris: OECD Publ.
Miflin, B. 1., Cullimore, 1. V., 1984: Nitrogen assimilation in the legume-Rhizobium
symbiosis. A joint endeavour. In: Verma, D. P. S., Hohn, Th. (eds.), Genes
Involved in the Microbe-Plant Interaction, Plant Gene Research, Vol. I, pp.
129-174. Wien - New York: Springer.
Minchin, F. R., Witty, 1. F., Sheehy, 1. E., Muller, M., 1983: A major error in the
acetylene reduction assay. Decreases in nodular nitrogenase activity under
assay conditions. 1. Exp. Bot. 34, 641-649.
Misaghi, I. 1., 1982: Physiology and biochemistry of plant-pathogen interaction.
New York: Plenum Press.
Mohapatra, S. S., Gresshoff, P. M., 1984: Sensitivity to oxygen of nitrogenase
activity in Rhizobium strain ANU289 of the non-legume Parasponia (Ulmaceae).
Aust. 1. BioI. Sci. 37, 31-36.
Mohapatra, S. S., PUhler, A., 1986: Detection of nodule specific polypeptides from
effective and ineffective root nodules of Medicago sativa. 1. Plant Physiol., sub-
mitted.
Mulligan, 1. T., Long, S. R., 1985: Induction of Rhizobium meliloti node by plant
exudate requires nodD. Proc. Nat. Acad. Sci. U.S.A. 82, 6609-6613.
Mulvaney, C. S., Hageman, R. H., 1984: Acetaldehyde oxime, a product formed
during the in vivo nitrate reductase assay of soybean leaves. Plant Physiol. 76,
118-124.
Milller, P., Niehaus, K., PUhler, A., 1985: Isolation and characterisation of mutants
of Rhizobium meliloti obtained by transposon Tn5 mutagenesis. In: Evans, H. 1.,
Bottomley, P. 1., Newton, W. E. (eds.), Nitrogen Fixation Research Progress,
pp. 137. Dordrecht: Nijhoff Publ.
Nadler, K. D., Avissar, Y. 1., 1977: Heme biosynthesis in soybean root nodules:
The role of bacterial gamma-aminolevulinic acid synthase and aminolevulinic
acid dehydrase in the synthesis of the heme of leghemoglobin. Plant Physiol.

60, 433-436.
Nambiar, P. T. c., Nigam, S. N., Dart, P. 1., Gibbons, R. W., 1983: Absence of root
hairs in non-nodulating groundnut, Arachis hypogaea. (L). 1. Exp. Bot. 34,
484-488.
Nelson, R L., Bernard, R L., 1984: Production and performance of hybrid soy-
beans. Crop Sci. 24, 549-553.
Nelson, R S., Ryan, S. A, Harper, 1. E., 1983: Soybean mutants lacking consti-
tutive nitrate reductase I. Selection and initial plant characterization. Plant
Physiol. 72, 503-509.
Nutman, P. S., 1946: Genetical factors concerned in the symbiosis of clover and
nodule bacteria. Nature 157,463-465.
Nutman, P. S., 1949: Physiological studies on nodule function II. The influence of
delayed nodulation on the rate of nodulation in red clover. Ann. Bot. 13,
261-283.
Nutman, P. S., 1953: Symbiotic effectiveness in nodulated red clover 1. Variation in
host and bacteria. Heredity 8, 35-46.
Nutman, P. S., 1969: Genetics of symbiosis and nitrogen fixation in legumes. Proc.
R Soc. London, Ser. B 172,417-437.
Ohlendorf, H., 1983 a: Selektion auf Resistenz von Pisum sativum gegen Rhizobium-
Stamm 311 d. Z. Pflanzenziicht. 90, 204- 221.
Ohlendorf, H., 1983 b: Untersuchungen zur Vererbung der Resistenz von Pisum
sativum gegen Rhizobium-leguminosarum-Stamm 311d. Z. Pflanzenziicht. 91,
13-24.
Olson, E. R, Sadowsky, M. 1., Verma, D. P. S., 1985: Identification of genes
involved in the Rhizobium-legume symbiosis by D1-Mu (kan:lac)-generated
transcript lesions. Biotechnology 3, 143-149.
Pacovsky, R. S., Bayne, H. G., Bethlenfalvay, G. 1., 1984: Symbiotic interactions
between strain of Rhizobium phaseoli and cultivars of Phaseolus vulgaris. Crop
Sci. 24, 101-105.
Pankhurst, C. E., Sprent, 1. I., 1975 a: Surface features of soybean root nodules.
Protop1asma 85, 85-98.
Pankhurst, C. E., Sprent, 1. I., 1975 b: Effects of water stress on the respiration and
nitrogen fixation ability of soybean root nodules. 1. Exp. Bot. 26, 287 -304.
Peterson, M. A, Barnes, D. K., 1981: Inheritance of ineffective nodulation and
non-nodulation traits in alfalfa. Crop Sci. 21, 611-616.
Phillips, D. A, Torrey, 1. G., 1970: Cytokinin production by Rhizobiumjaponicum.
Physiol. Plant 23, 1057 -1063.
Phillips, D. A, Bedmar, E. 1., Qua1set, C. 0., Teuben, L. R, 1985: Host legume
control of Rhizobium function. In: Ludden, P. W., Burris, 1. E. (eds.), Nitrogen
Fixation and CO 2 Metabolism, pp. 203-212. Amsterdam: Elsevier.
Pierce, M., Bauer, W. D., 1983: A rapid regulatory response governing nodulation
in soybean. Physiol. Plant 73, 286-290.
Price, G. D., Mohapatra, S. S., Gresshoff, P. M., 1984: Structure of nodules formed
by Rhizobium strain ANU289 in the non-legume Parasponia and the legume
siratro (Macroptilium atropurpureum). Bot. Gaz. 145,444-451.
Ralston, E. 1., Imsande, 1., 1982: Entry of oxygen and nitrogen into intact nodules
of soybean. 1. Exp. Bot. 33, 208-214.
Rennie, R 1., Dubetz, S., 1984: Multistrain versus single strain Rhizobium
japonicum inoculants for early maturing (00 and 000) soybean cultivars: N2
fixation quantified by 15N isotope dilution. Agron. 1. 76, 498-502.
Rennie, R J., Kemp, G. A., 1984: 15N determined time course for nitrogen fixation
in two cultivars of field bean. Agron. J. 76, 146-154.
Robertson, J. G., Wells, B., Bisseling, T., Farnden, K. J. F., Johnston, A W. B.,
1984: Immunogold localisation of leghaemoglobin in the plant cytoplasm in
nitrogen fixing root nodules of pea. Nature 311, 254-256.
Rolfe, B. G., Shine, J., 1984: Rhizobium-Leguminosae symbiosis: The bacterial point
of view. In: Verma, D. P. S., Hohn, Th. (eds.), Genes Involved in Microbe-plant
Interaction, Plant Gene Research, Vol. 1, pp. 95-128. Wien - New York:
Springer.
Ronson, C. W., Astwood, P. M., Downie, J. A, 1984: Molecular cloning and genetic
organisation of C 4 -dicarboxylate transport genes from Rhizobium legumin-
osarum. J. Bacteriol. 160, 903-909.
Ronson, C. W., Lyttleton, P., Robertson, J. G., 1981: C 4 -dicarboxylate transport
mutants of Rhizobium trifolii from ineffective nodules of Trifolium repens. Proc.
Nat. Acad. Sci. U.S.A 78, 4284-4288.
Ryan, S. A., Nelson, R S., Harper, J. E., 1983: Soybean mutants lacking consti-
tutive nitrate reductase activity II. Nitrate assimilation, chlorate resistance and
inheritance. Plant Physiol. 72, 510-514.
Sandeman, R. A., Gresshoff, P. M., 1985: Nitrogenase activity and inactivation in
isolated bacteroids from the legume siratro and the non-legume Parasponia
rigida. Plant Sci. Lett. 37, 199-204.
Schmidt, J., John, M., Kondorosi, E., Kondorosi, A., Wieneke, U., Schroder, G.,
Schroder, J., Schell, J., 1984: Mapping of the protein encoding region of the
Rhizobium meliloti common nodulation genes. EMBO J. 3, 1705-1711.
Schofield, P. R., Djordjevic, M. A, Rolfe, B. G., Shine, J., Watson, J., 1983: A
molecular linkage map of the nitrogenase and nodulation genes of Rhizobium
trifolii. Mol. Gen. Genet. 192, 456-466.
Schofield, P. R., Ridge, R W., Rolfe, B. G., Shine, J., Watson, J., 1984: Host spe-
cific nodulation is encoded on a 14 Kb DNA fragment in Rhizobium trifolii.
Plant Mol. BioI. 3, 3-15.
Schubert, K., 1981: Enzymes of purine biosynthesis and catabolism in Glycine max.
Plant Physiol. 68,1115-1122.
Schubert, K., Evans, H. J., 1976: Hydrogen evolution: a major factor affecting the
efficiency of nitrogen fixation in nodulated soybeans. Proc. Nat. Acad. Sci.
U.S.A. 73, 1207-1211.
Schuller, K. A., Day, D. A., Gibson, A. H., Gresshoff, P. M., 1986: Effect of nitrate
on nitrogen fixation and ammonia assimilation in soybean nodules. Plant
Physiol. 80, in press.
Scott, K. F., Rolfe, B. G., Shine, J., 1983: Biological nitrogen fixation: primary
structure of the Rhizobium trifolii iron-protein gene. DNA 2, 149-158.
Sen, D., Weaver, R W., 1984: A basis for different rates of nitrogen fixation by the
same strains of Rhizobium in peanut and cowpea nodules. Plant Sci. Lett. 34,
239-246.
Sequira, L., 1984: Recognition systems in plant-pathogen interactions. BioI. Cell 51,
281-286.
Sheehy, J. E., Fishbeck, K. A, Dejong, T. M., Williams, L. E. Phillips, D. A, 1980:
Carbon exchange rates of shoots required to utilize available acetylene
reduction capacity of soybean and alfalfa root nodules. Plant Physiol. 66,
101-104.
Sheehy, J. E., Minchin, F. R, Witty, J. F., 1983: Biological control of the resistance
to oxygen flux in nodules. Ann. Bot. 52, 565-572.
Sinclair, T. R., Goudriaan, J., 1981: Physical and morphological constraints on

transport in nodules. Plant Physiol. 63, 143-145.
Singleton, P. W., Stockinger, K. R., 1983: Compensation against ineffective nodu-
lation in soybean. Crop Sci. 23, 69-72.
Skot, L., 1983: Cultivar and Rhizobium strain effects on the symbiotic performance
of pea (Pisum sativum). Plant Physiol. 59, 585-589.
Smartt, J., 1984: Gene pools of grain legume. Econ. Bot. 38, 24-35.
Smith, G. R., Knight, W. E., 1984: Inheritance of ineffective nodulation in crimson
clover. Crop Sci. 24, 601-604.
Spreit, L., Nelson, R. S., Harper, J. E., 1985: Nitrate reductases from wild type and
nrl-soybean (Glycine max (L) Merr.) leaves I. Purification, kinetics and physical
properties. Plant Physiol. 78, 80-85.
Streeter, J. G., 1977: Asparaginase and asparagine transaminase in soybean leaves
and root nodules. Plant Physiol. 60, 235-239.
Summons, R. E., Letham, D. S., Gollnow, B. I., Parker, C. W., Entsch, B., Johnston,
L. P., MacLeod, J. K., Rolfe, B. G., 1981: Cytokinin translocation and meta-
bolism in species of Leguminosae: studies in relation to shoot and nodule
development. In: Guem, J., Peaud-Lenoel, C. (eds.), Metabolism and
Molecular Activity of Cytokinins, pp. 69-79. New York - Berlin - Heidelberg:
Springer
Sutton, W. D., Paterson, A. D., 1983: Further evidence for plant host effect on Rhi-
zobium bacteroid viability. Plant Sci. Lett. 30, 33-41.
Tanner, J. W., Anderson, I. C., 1963: Investigation on non-nodulating soybean
strains. Can. J. Plant Sci. 43, 542-546.
Thimann, K. V., 1936: On the physiology of the formation of nodules on legume
roots. Proc. Nat. Acad. Sci. U.S.A. 22, 511-514.
Thomas, R. J., Jokinen, K., Schrader, L. E., 1983: Effect of Rhizobium japonicum
mutants with enhanced nitrogen fixation activity on nitrogen transport and
photosynthesis of soybean during vegetative growth. Crop Sci. 23, 453-456.
Tjepkema, J. D., Yocum, C. S., 1974: Measurement of oxygen partial pressure
within soybean nodules by microelectrode. Planta 119, 351-360.
Trinick, M. J. 1973: Symbiosis between Rhizobium and the non-legume Trema
aspera. Nature (London) 244, 459-460.
Tran, T. V. K., Toabart, P., Cousson, A., Darvil, A. G., Gollin, D. J., Chelf, P.,
Albersheim, P., 1985: Manipulation of the morphogenetic pathways of tobacco
explants by oligosaccharides. Nature 314, 615-617.
Tsein, H. C., Dreyfus, B. L., Schmidt, E. L., 1983: Initial stages in the morpho-
genesis of nitrogen fixing stem nodules of Sesbania rostrata. J. Bacteriol. 156,
888-897.
Tudge, C., 1984: Whatever happens to nitrogen? New Scientist 9, 13-15.
Van den Bos, R., Schots, A., Hentelez, J., van Kammen, A., 1983: Constitutive
nitrogenase synthesis from de novo transcribed mRNA in isolated Rhizobium
leguminosarum bacteroids. Bioch. Biophys. Acta 740, 313-322.
Vance, C. P., Johnson, L. E. B., 1983: Plant induced ineffective nodules in alfalfa
(Medicago sativa L): structural and biochemical comparisons. Can. J. Bot. 61,
93-106.
Verma, D. P. S., Nadler, K. D., 1984: Legume-Rhizobium symbiosis: The host's
point of view. In: Verma, D. P. S., Hohn, Th. (eds.) Genes Involved in Microbe-
Plant Interaction, Plant Gene Research, Vol. 1, pp. 58-84. Wien - New York:
Springer.
Verma, D. P. S., Lee, J. S., Katinakis, P., Sutton, B., 1985: Nodule specific genes of
soybean. In: Analysis of Plant Genes Involved in the Legume-Rhizobium Sym-

biosis, pp. 74-84. Paris: OECD Pub!.
Vest, G., 1970: Rj3 - a gene controlling ineffective nodulation in soybean. Crop
Sci. 10, 34-35.
Vest, G., Caldwell, B. E., 1972: Rj4 - a gene controlling ineffective nodulation in
soybean. Crop Sci. 12, 692.
Viands, D. R., Vance, C. P., Heichel, G. H., Barnes, D. K., 1979: An ineffective
nitrogen fixation trait in alfalfa. Crop Sci. 19,905-908.
Weber, C. R., 1966: Nodulating and non-nodulating soybean isolines II. Response
to applied nitrogen and modified soil conditions. Agron. J. 58, 46-49.
Weinman, J. J., Fellows, F. F., Gresshoff, P. M., Shine, J., Scott, K. F., 1984: Struc-
tural analysis of the genes encoding the molybdenum-iron protein of
nitrogenase in the Parasponia Rhizobium strain ANU289. Nuc. Acids Res. 12,
8329-8344.
Werner, D., Morschel, E., Kort, R., Mellor, R. B., Bassarab, S., 1984: Lysis ofbacte-
roids in the vicinity of the host cell nucleus in an ineffective (frx- ) root nodule
of soybean (Glycine max). Planta 162, 8-16.
Werner, D., Christensen, T., Mellor, R. B., Morschel, E., 1985: Glycosyltransferases.
and the peribacteroid membrane. In: Analysis of the Plant Genes Involved in
the Legume- Rhizobium Symbiosis, pp. 61-71. Paris: 0 ECD Pub!.
Whitmore-Smith, D., 1985: Studies on the constitutive nitrate reductase activity of
different soybean mutants. Honours dissertation. Botany Department, Aus-
tralian National University, Canberra, Australia.
Williams, L. F., Lynch, D. L., 1954: Inheritance of a non-nodulation character in
soybean. Agron. J. 46, 28-29.
Williams, L. E., Dejong, T. M., Phillips, D. A., 1982: Effect of changes in shoot
carbon exchange rate on soybean nodule activity. Plant Physio!. 69, 432-436.
Winarno, R., Lie, T. A., 1979: Competition between Rhizobium strains in nodule
formation. Interaction between nodulating and non-nodulating strains. Plant
and Soil 51, 135-142.
Witty, J. F., Minchin, F. R., Sheehy, J. E., Minguez, M. I., 1984: Acetylene-induced
changes in the oxygen diffusion resistance and nitrogenase activity of legume
root nodules. Ann. Bot. 53, 13-20.
Witty, J. F., Skot, L., Revsbech, N. P., 1985: Direct evidence for a variable barrier to
oxygen diffusion into legume nodules. In: Evans, H. J., Bottomley, P. J.,
Newton, W. E. (eds.), Nitrogen Fixation Research Progress, p. 355. Dordrecht:
Nijhoff Pub!.
Chapter 7
Endosperm Proteins
Peter I. Payne
Plant Breeding Institute, Maris Lane, Trumpington, Cambridge CB 2 2 LQ, U.K.
With 7 Figures
Contents
I. Introduction
II. Origin and Development of the Endosperm
III. Classification of the Major Endosperm Proteins
IV. Biochemical Complexity and Genetic Variation of Endosperm Proteins
V. Gene Mutations
A. Role in Plant Breeding
B. Role in Genetics
C. Role in Biochemistry and Molecular Biology
D. Role in the Food Industry
VI. Chromosome Mutations
VII. Conclusions
VIII. References
I. Introdnction
Nearly all the major crop plants of the world are cereals, comprising in
decreasing order of production, wheat, maize, rice, barley, sorghum, oats,
millet and rye (Harlan and Starks, 1980). The major organ by volume of the
cereal grain is the endosperm which serves virtually exclusively as a store
of food reserves for the germinating seedling.
The major macromolecule in the endosperm of all cereals is starch and
it amounts to some 80-90 % of the total dry weight. The next most
abundant macromolecule is protein. Its concentration varies according to
the cereal and to the method of cultivation but is generally highest in wheat
and oats, with a usual range of 10-17 %, and lowest in maize and rice,
when it can fall to 6 % (Altschul, 1965). The concentration of protein in the
cereal endosperm is appreciably lower than in leguminous seeds, where the
storage organ is not the endosperm but the cotyledon and the range of
208 P. I. Payne
protein content is usually from 20-40 % of the dry weight (Altschul, 1965).
Nevertheless, because of the relative production of cereals and legumes,
the decreasing order of protein production amongst these crops is wheat,
maize, soybean, rice, barley, oats, sorghum, peanut, millet, rye, peas and
beans (Harlan and Starks, 1980). The majority of the peoples of the world,
therefore, obtain most of their protein by eating cereals and it is not sur-
prising that protein amount and protein quality feature strongly in many
cereal breeding programmes and that their biochemistry and genetics are
being studied actively.
The object of this chapter is to show how mutations involving endo-
sperm protein structural genes or genes which control their output are
being exploited, not only by the plant breeder but also by the plant
biochemist and the molecular biologist as a means to understanding the
structure, synthesis and regulation of these proteins.
II. Origin and Development of the Endosperm
The endosperm originates at the time of fertilisation in the embryo sac of

the ovum by a two-step process. First, the two haploid, polar nuclei of the
embryo sac coalesce to form the diploid, fusion nucleus. This in turn fuses
with one of the two sperm nuclei derived from a single pollen grain to form
the triploid endosperm nucleus. The remaining sperm nucleus fuses with
the egg-cell nucleus to form a diploid zygote which gives rise to the
embryo. The endosperm's triploidy has important consequences in bio-
chemical and genetic analyses.
With the exception of a few species such as orchids, the endosperm
tissue grows rapidly during the early stages of seed development. It
becomes senescent in most plant species well before seed maturity, is
rapidly digested and is used as a food source for the developing embryo.
But in other species, and particularly in all the cereals, the endosperm con-
tinues to develop and becomes the primary food store of the mature seed.
The formation of the endosperm and its development during seed matu-
ration have been described in detail for wheat by Simmonds and O'Brien
(1981). In the remainder of this section, I will restrict myself to a very brief
description of the synthesis and accumulation of protein in cereal endo-
sperm, with the emphasis on wheat.
Storage protein, to be discussed and defined in detail in the next
section, accumulates in the cell expansion phase of endosperm devel-
opment. For wheat, this is from about 12 to 40 days after anthesis on a time
scale of 50 days from anthesis to grain maturity (Parker, 1980). The
molecular events in the synthesis of storage protein have been elucidated
and they resemble the synthesis of protein in animal cells which are des-
tined for export. The messenger RNAs encoding storage proteins pass
from nucleus to cytoplasm and combine with ribosomes to form poly-
somes. Protein synthesis commences with the formation of a signal
sequence at the N-terminus. This enables these polysomes to specifically
Endosperm Proteins 209
attach to the endoplasmic reticulum. As synthesis of the protein chain con-

tinues, the leader sequence and the rest of the protein pass into the lumen
of the endoplasmic reticulum, where the signal sequence is removed
(barley: Cameron-Mills et a/., 1978; maize: Burr and Burr, 1981).
From the endoplasmic reticulum the proteins pass into storage organ-
elles, the protein bodies, but the mechanism of the movement is disputed in
some cereals. In maize the mechanism is clear: storage protein accumulates
in localised regions of the endoplasmic reticulum which swell and even-
tually bud off as membrane-bound protein bodies (Larkins and Hurkman,
1978). The majority of work in barley and wheat (e. g. Cameron-Mills and
von Wettstein, 1980; Parker, 1982, respectively) has implicated the
involvement of Golgi bodies. For these species it is suggested that storage
proteins pass to the Golgi bodies, with which the endoplasmic reticulum is
directly connected. The protein is then packaged into vesicles and trans-
ferred to vacuoles which develop into protein bodies. This mechanism is
very similar to that accepted for storage protein accumulation in legumes.
However, Shewry and Miflin (1985) believe that storage protein accumu-
lates in barley and wheat as it does in maize without the involvement of
Golgi. Very recently, Parker (unpublished observations) has presented evi-
dence that both mechanisms exist, with endoplasmic-reticulum budding
being prevalent in those cells in the centre of the endosperm which accu-
mulate least protein.
The major function of the storage proteins of the endosperm is to serve
the embryo as a supply of elements, particularly nitrogen, in the early
stages of germination prior to the seedling becoming established. Since a
specific amino-acid sequence is not essential to fulfil the role of these pro-
teins, change through mutation will be tolerated more than changes in the
sequence of enzymic proteins or structural proteins in multi-molecular
complexes such as ribosomes. Nevertheless, some constraints on storage-
protein amino-acid sequences must exist. Other than the maintenance of
specific sequences in the structural genes which are essential for the pro-
duction of all proteins, the leader sequence must be preserved and all parts
of protein structure involved in the transport and packaging of the pro-
teins. For maximum efficiency, storage proteins contain large percentages
of nitrogen-rich amino-acid residues (asparagine or glutamine) and, to aid
packing, high percentages of hydrophobic amino-acid residues such as
proline, leucine and isoleucine and low percentages of hydrophylic amino
-acid residues such as lysine, glutamic and aspartic acids. Major changes in
the balance of these amino acids are unlikely to be tolerated. The rate of
mutation of different storage-protein groups will be discussed in a later
section of this chapter.
III. Classification of the Major Endosperm Proteins

The classification of the endosperm proteins of cereals has been dominated
by Osborne (1907) and his contemporaries and is based on the proteins'
relative solubilities in different solvents:
210 P. I. Payne
1. albumin: soluble in water;

2. globulin: soluble in salt solutions but insoluble in water;
3. prolamin: soluble in aqueous alcohols;
4. glutelin: soluble in dilute acids and alkalis.
For the prolamin group of proteins, specific names have been given for
each cereal. Thus the prolamin of maize is zein, that of barley is hordein,
that of wheat is gliadin and that of oats avenin. In wheat the glutelin
fraction is usually called glutenin. From the above definitions, wheat and
sorghum were considered to contain about equal amounts of prolamin and
glutelin, maize and barley rather more prolamin than glutelin, whereas rice
was thought to be dominanted by glutelin and oats by globulin.
In 1982, Payne and Rhodes stated: "These terms for the different
groups of proteins are still used today although modern workers use chem-
icals and extraction procedures which are different from those used at the
beginning of the century. This resulted in changed meanings for these
protein groups and unfortunately there are no simple, modern definitions
that are acceptable to all. The major problems lie with prolamin and
glutenin. "
Fortunately the last two or three years have seen rapid advances in our
understanding of these proteins from all the major cereals. This has been
due partly to N-terminal amino-acid sequencing of isolated proteins but
mainly to DNA sequencing of storage-protein genes cloned in bacterial
vectors. Our present understanding of these storage proteins has been
covered very recently by an extensive review (Shewry and Miflin, 1985).
In wheat, the sequencing of storage-protein genes is well advanced.
Extensive sequences for a-gliadins, y-gliadins and low-molecular-weight
(LMW) subunits of glutenin have been published (Bartels and Thompson,
1983; Kasarda et ai., 1984; Rafalski et al., 1984). From this work it is likely
that the genes of all the protein groups above originated from the dupli-
cation of a common ancestral gene. Thus, with the possible exception of
high-molecular-weight (HMW) glutenin subunits whose sequence appears
to be basically unrelated (Thompson et al., 1983) there is only one major
class of storage protein in wheat endosperm and this would most appropri-
ately be called prolamin. Within the prolamin group are proteins with pro-
lamin-type and glutenin-type solubility in spite of the similarity of their
primary sequences. This is because glutenin subunits occur as disulphide
bonded aggregates which greatly reduces their solubility, whereas gliadins
occur as simple, monomeric molecules and contain only intramolecular
disulphide bonds. The distinction between gliadin and glutenin is certain
to continue to be used because the two groups have different functional-
ities in food processing, particularly in bread making, but it should be re-
membered that both lie within the prolamin subgroup of proteins.
A further complication in the old classification of proteins by solubility
is that some proteins in the glutenin fraction are not true storage proteins
but are probably structural proteins of macromolecular complexes such as
ribosomes or membranes. These therefore are not part of the modern pro-
lamin fraction and they have not been studied extensively.
Sequencing has demonstrated a close evolutionary relationship

between the prolamins of barley, rye and wheat (Shewry et ai., 1980; Miflin
et ai., 1984) which are all members of the subfamily, Triticeae. The other
cereals which contain mainly prolamins are maize, sorghum and millet,
three cereals which are closely related to each other but distant from
barley, rye and wheat. This evolutionary distance is reflected in the very
different prolamin amino-acid sequences of maize (Hu et ai., 1982) and the
three cereals of the Triticeae. The two types of prolamins nevertheless have
similar amino-acid compositions being rich in glutamine and hydrophobic
amino acids and deficient in acidic and basic amino-acid residues such as
lysine. Prolamin molecules are characteristic in containing repeated
sequences each several amino-acid residues long.
Oat is unique amongst cereals in containing protein which is mainly
soluble in salt solutions. Peterson (1978) demonstrated that the protein
occurs as a large protein, 12 S with a molecular weight of 322,000. He gave
evidence that it is built up from six subunits of molecular weight 21,700
and six subunits of 31,700. This and other evidence shows that the protein
resembles the 12 S, legumin group of globulin storage proteins found in
leguminous seeds.
A particularly exciting development recently has been the finding that
at least some of the major components of rice glutenin have a subunit
structure which resembles that of oat and legume 12 S globulin (Yamagata
et ai., 1982). Furthermore, amino-acid sequencing of one component has
demonstrated a 25 % homology with a subunit of pea legumin (Zhae et aI.,
1983). Presumably this rice globulin must have undergone mutational
change during evolution, resulting in a greater proportion of hydrophobic
amino-acid residues in the protein and causing it no longer to be soluble in
salt solutions.
For the purposes of this chapter, we can conclude in the light of recent
research that cereals can contain two major storage protein groups, glob-
ulins and prolamins, the former being related to counterparts in legu-
minous seeds.
IV. Biochemical Complexity and Genetic Variation of Endosperm Proteins
During the course of evolution the prolamin genes in particular have dupli-
cated and undergone mutational change to form families of different but
related genes. These families have, during the course of time, become par-
tially split up by chromosomal rearrangements. The outcome has resulted
(1) in individual cereal varieties containing complex mixtures of endo-
sperm proteins and (2) significant variation occuring between varieties for
protein type.
In Fig. 1 the endosperm proteins of a variety of bread wheat have been
fractionated by two-dimensional electrophoresis. In total there are about
60 major components and about double that number of minor components.
212
NEPHGE
P. I. Payne
IEF
,
w
C!J
~
I
CIJ
C
CIJ
Fig. 1. Two-dimensional fractionation of the endosperm proteins of the wheat

variety Sicco. There are two first dimensions, IEF (isoe1ectric focussing) to frac-
tionate the acidic proteins and NEPHGE (non-equilibrium, pH gradient electro-
phoresis) to resolve the more basic proteins. The two separations were fractionated
side-by-side on the same second-dimension gel of SDS-PAGE. Further details of
the method of grouping the proteins into glutenin subunits, gliadins and non-
storage groups are given in Payne et al. (1984 b)
The complexity in wheat has been exaggerated as it possesses three dif-

ferent genomes, the storage proteins from each genome being distin-
guishable. For comparison, a separation of a variety of diploid barley is
also shown (Fig. 2). The complexity is less, though at least 40 major compo-
nents have resolved. A similar order of complexity to that in barley has
been shown to occur in maize (Burr and Burr, 1981).
In Figs. 3 and 4 the proteins of different varieties of wheat and barley,
respectively, have been fractionated by one-dimensional electrophoresis to
display the great variation, presumably due to the presence of different
alleles at each storage-protein gene locus. Interestingly, the variation re-
vealed by the same technique is very much less for millet, sorghum (Fig. 5)
NEPGE IEF
w
~
~
I
en
C
en
Fig. 2. Two-dimensional fractionation of the endosperm proteins of barley variety,

Sundance. The electrophoretic procedure is described in the legend to Fig. 1
and maize. The technique, SDS-PAGE, fractionates mainly according to

molecular weight. Variation in the prolamins of the above three cereals is
much more by charge than molecular weight and such techniques as iso-
electric focussing do show extensive allelic variation (Rhighetti et ai.,
1977).
The globulin storage proteins of cereals show much less variation
amongst varieties than do prolamins. Thus, amongst nine varieties of oats,
there was very little variation in the globulin fraction when assessed by
SDS-PAGE and only slightly more when fractionated by IEF (Robert et
al., 1983 a). When the prolamins (avenins) from the same set of varieties
were analysed by SDS-PAGE and IEF, variation was much more prom-
inent (Robert et ai., 1983 b). Similarly there is very little electrophoretic var-
iation amongst globulins (glutelins) from varieties of rice (Park and
Stegemann, 1979).
The likely interpretation of these contrasting results with globulins and
prolamins is that the latter are undergoing more rapid evolutionary change
that the former. This is probably because the globulins have a far more
elaborate packaging system in the protein bodies. Globulin polypeptides
are arranged into an ordered structure of an approximate molecular weight
of 300,000 (Peterson, 1978) imparting sequence constraints to conserve
214 P. I. Payne
____ __ _ Origin
HMW {
subunits
of ~92 , 500
glutenin
~68 , OOO
~39 , OOO
Fig. 3. Fractionation of the total endosperm proteins of 20 varieties of wheat by

SDS-PAGE (sodium dodecyl sulphate, polyacrylamide-gel electrophoresis). The
arrows indicate molecular weights obtained from the fractionation of standard pro-
teins
molecular shape. For prolamins, which in the main have a much more
hydrophobic nature, deposition in protein bodies is probably only by prec-
ipitation.
V. Gene Mutations
The variation in cereal endosperm proteins amongst varieties, described in

Section IV, has occured over the millenia as a result of the mutation of struc-
tural genes and probably also of genes regulating endosperm protein syn-
thesis. Further variation in the regulatory genes has been induced either with
chemicals or by irradiation. Mutations in single structural genes have in
general very little effect in changing the composition of grain storage pro-
teins because the genes occur in nearly all cases in large families. However,
with chromosome mutations (see Section IV) effects can be marked.
--
- 9 4,000
- 6 8,000
_43,000
_30 ,000
Fig. 4. Fractionation of endosperm proteins of nine varieties of barley by

SDS-PAGE
A. Role in Plant Breeding

The cereals contain a small percentage of protein compared with legume
seeds and those which store predominantly prolamins have in particular a
poor balance of amino acids for human nutrition. The first limiting amino
acid is invariably lysine (Altschul, 1965). Much effort has therefore gone
into screening for higher protein and high lysine content in germplasm col-
lections and in lines treated with mutagens. Success was achieved, rather
easily for protein content, and this brought a flurry of activity throughout
the world by cereal breeders to bring these so-called "high-protein" and
"high-lysine" genes into elite, commercial varieties. The results have so far
been disappointing, in that few if any novel varieties with higher-than-
normal protein or lysine have been developed which have outclassed and
replaced conventional varieties in agriculture. Some of the problems have
included low or erratic grain yield, reduced protein yield, small grain size
and increased susceptibility to certain diseases (IAEA, 1984). Nevertheless
in several major breeding institutes, continued and detailed work is
steadily reaping rewards. At CIMMYT for instance the opaque-2 genes of
maize, which confer the high lysine trait, have been transferred from unde-
sirable floury endosperm types to a vitreous endosperm type which is hard
216 P. I. Payne
_66,000
_24 ,000
_14 , 300
1 5 6 11
Fig. 5. Fractionation of grain protein of five sorghum varieties (slots 1-5) and six
millet varieties (slots 6-11) by SDS-PAGE
milling and new lines are beginning to compare favourably with com-
mercial varieties for yield and protein content (Vasal et al., 1984). Transfer
of high protein genes from emmer (Triticum dicoccoides) to bread-wheat
also looks promising (e. g. Grama et al., 1984). Indeed in the most recent
proceedings of a conference on cereal grain improvement (IAEA, 1984),
various groups working on maize, barley and wheat are still optimistic
about transferring high-protein and high-lysine genes into commercially
acceptable varieties.
The great majority of high-lysine genes have so far been detected in
barley and maize as described below in Section V.C, but none have been
found in wheat. Vogel et al. (1973) screened some 20,000 entries in the
world wheat collection of the United States Department of Agriculture and
found only 0.5 % of genetic variation in total grain protein content was due
to lysine. A likely reason for this failure is probably that wheat is hexa-
ploid: a mutant gene conferring high lysine would only occur in one of the
three genomes and so might not be detected.
Nevertheless, the endosperm proteins of wheat are becoming more

important in the breeding of new varieties, not with respect to improved
nutritional quality, but to improved bread-making quality. In wheat-
growing areas of the world where water is plentiful and the growing season
is long, as in Western Europe, grain yield tends to be high but protein
content low (11 % of dry weight or lower) even with the application of
nitrogen fertilizers to attain optimal yield. Low protein content has an
adverse affect on bread-making quality and if varieties are to be grown in
these areas for bread production they must have a good protein quality.
Variation in protein quality is primarily caused by the great variation in
protein components which occurs between genotypes (see Fig. 3). Gliadins
are monomeric proteins which give extensibility to a dough whereas glut-
enins consist of approximately 16 different subunits (Payne and Corfield,
1979) disulphide-bonded into large, elongated fibrils which impart strength
and elasticity (Wall, 1979; Payne et al., 1984 b). It is generally agreed that it
is the elastic component which is often limiting for bread-making (Orth
and Bushuk, 1972). Payne et al. (1981) were the first to demonstrate
convincingly, by the analysis of segregating progeny, that the different sub-
units which make up glutenin have different associations with good bread-
making quality. By the production of more accurately defined genotypes
these associations are being studied in more detail (Payne et al., 1984 b).
Glutenin subunit genes occur at six different major loci on three of the
21 chromosomes of bread-wheat. The genes at each locus display extensive
allelic variation and, for several loci, different alleles are associated with
different degrees of elasticity in native glutenin (Payne et al., 1984 b). Cur-
rently, wheat breeders at the Plant Breeding Institute, Cambridge, are
selecting parents in their crossing programmes which have complementary,
good-quality glutenin subunits (i. e. their genes occur at different loci) so
that they can be combined in a few progeny which will have better quality
than either parent. Gel electrophoresis is being used in the breeding pro-
gramme at fairly early generations as a secondary screen to select the few
progeny which have the ideal protein composition for a particular cross
(Payne et al., 1984b).
B. Role in Genetics
As will be described in Section VI, chromosome mutations have been
exploited by geneticists to locate the genes which code for storage proteins
to chromosomes. Gene mutations have been used to study the location and
distribution of the genes on individual chromosomes. In most cereals, the
prolamin genes occur as a series of tightly linked gene families at complex
loci. Genetic analysis of crosses has shown that the variation in protein
pattern which occurs between varieties is due to the presence of numerous
complex alleles at each locus.
In wheat, there are nine major complex loci and a few minor loci
coding for storage proteins, all of which occur on the homoeologous group
one and group six chromosomes. The relative positions of the loci on the
218 P. I. Payne
centiMorgans(cM)
I I I I I
20 0 20 40 60
1A
L s
/
L
I 0
I I s
I
18
/
L
I 0
I I s
I
10
I
I ()
I ~ I I
Glu-l Trp-l Glu-2 GIi-l
Fig. 6. Chromosomal location and position of the storage protein loci on the group
I chromosomes of bread wheat. The two major groups of loci are Glu-l and Gli-l.
The former code for HMW glutenin subunits and the latter for families of LMW
glutenin subunits, w-gliadins and y-gliadins. The Trp-l loci probably code for
minor globulin-type proteins (Singh and Shepherd, 1985) and the Glu-210ci minor
LMW subunits of glutenin. The position of Glu-2 on chromosome I D has not been
determined. The other main loci, Gli-2, occur on the short arms of chromosomes
6A, 6B and 6D and code for a- and ~-gliadins (L = long, S = short)
chromosomes was determined by making crosses between lines with con-

trasting storage-protein alleles and determining the frequency of recombi-
nation in the progeny. The chromosomal position of Glu-1, coding for
HMW subunits of glutenin (Fig. 6), appears to have been remarkably stable
during the course of evolution and is one of several stable markers which
are being used to understand chromosome homologies in the Triticeae. In
rye the equivalent locus, Glu-R 1, occurs on chromosome 1 R (formerly E)
and in barley the locus, called Hor 3, is on chromosome 5 (Shewry et al.,
1984 a; see Fig. 2). Both loci lie close to the centromeres on the long arm of
their respective chromosomes and also contain one or more prolamin loci
on the short arms of the same chromosomes which are at least partially
equivalent to Gli-1 of wheat (Shewry et al., 1984 a). It is therefore con-
cluded that chromosomes 1 A, 1 Band 1 D of wheat are homologous to
chromosomes 1 R of rye and 5 of barley. Simply from the chromosomal
location of other loci homologous to Glu-l, other homologous chromo-
somes are likely to be 1 U of Aegi/ops umbellulata, I of Ae. elongatum (Law-
rence and Shepherd, 1981) and G of Hordeum chilense (Payne, Holt and
Miller, unpublished results).
C. Role in Biochemistry and Molecular Biology ...

The so-called high-protein genes discussed in Section V.A probably vary
widely in their functions, for instance by causing improved nitrogen uptake
by roots or. nitrogen translocation from senescing leaves and stems, or by
changing plant height. In contrast the high-lysine genes directly affect
storage-protein accumulation and they are being actively exploited by
biochemists and molecular biologists to understand the synthesis and regu-

lation of prolamin and globulin in the developing endosperm.
In barley, several lines containing higher-than-normal lysine contents
have been described (reviewed by Miflin and Shewry, 1979) and three of
them are well characterised. These are Risq, 56, obtained from treating
variety Carlsberg II with y-rays, Risq, 1508, derived from ethyleneimine-
treated grain of Bomi, and Hiproly, an Ethiopian landrace naturally high
in lysine. All contain reduced levels of hordein but this is achieved by dif-
ferent mechanisms and to different degrees (Table 1).
Table 1. Properties and chromosome location of the commonly studied "high-

lysine" genes of cereals
Prolamin Synthesis
Chromo inhi- groups
Cereal Line Mutation Locus Expression
some bition sup-
% pressed
Barley Hiproly Spontaneous lys 7 Recessive All

Barley Ris<p 56 Induced Hor2ca 5 Co-dom- 25 B-hordein
inant
Barley Ris<p 1508 Induced lys3a 7 Recessive 69 All
Maize Opaque-2 Spontaneous O2 7 Recessive 47 Mainly
20,000
Maize Opaque-6 Spontaneous 06 Recessive 89 All
Maize Opaque-7 Spontaneous 07 10 Recessive 78 Mainly
20,000
Maize Floury-2 Spontaneous Fl2 4 Semi-dom- 35 All
inant
Data taken from Doll (1984) and Soave and Salamini (1984).
The high lysine content of Risq, 56 was shown by Doll (1980) to be due to
the presence of a non-functional allele, Hor 2 ca, at the Hor 2 locus which
codes for B hordeins, one of two major prolamin groups in barley. By SDS-
PAGE, it was shown that the synthesis of the major B hordeins was com-
pletely suppressed. Several minor bands were still present but it was not
determined whether their genes actually occur at Hor 2. Recently, by
methods in molecular biology, Kries et al. (1983) showed that in Risq, 56,
the Hor 2 locus was partially or completely deleted. It is very likely,
therefore, that Risq, 56 has arisen from a chromosome mutation which
resulted in the loss of a very small, interstitial segment of chromosome. In
this mutant there is a doubling in the production of C hordein (Koie and
Doll, 1979), the other major prolamin, compared to the parent variety
Carlsberg II, and a great increase in four salt-soluble proteins; protein z, ~
amylase and chymotrypsin inhibitors CI-l and CI-2. The salt-soluble pro-
teins are responsible for raising the lysine content of the grains for they
contain between 5 and 11 % lysine.
220 P. I. Payne
Unlike the Hor2ca locus of Risq, 56, the lys3a locus of Risq, 1508 and
the lys locus of Hiproly are not alleles of prolamin genes and they also
occur on different chromosomes (Table 1). The Risq, mutant 1508 shows a
drastic reduction in hordein accumulation and has a correspondingly high
lysine content (Doll, 1984). Hiproly itself also contains a much reduced
hordein content but when crossed and backcrossed to commercial varieties,
selecting for the lys locus at each generation, the hordein content is only
slightly lower than that of the recurrent parent. This led Tallberg (1984) to
conclude that the lys gene has little direct effect on hordein accumulation.
The two genes, lys and lys 3 a, also have different effects on the rate of syn-
thesis of the lysine-rich proteins: in Hiproly, protein z, ~-amylase and
chymotrypsin inhibitors CI-l and CI-2 are over-produced as in Risq, 56
(Doll, 1984). In Risq, 1508 however, CI-l is preferentially increased together
with free lysine but ~-amylase is strongly inhibited (Doll, 1983). Clearly a
comparative analysis of the lys and lys 3 a genes at the molecular level may
give insight into the derepression and expression of the hordein genes.
Research in this area has barely started but Kries et al. (1984) have indicated
that the effect of lys 3 a on hordein synthesis is either at the level of tran-
scription or on the early processing of messenger RNA.
In maize, like barley, several high-lysine lines have been developed and
described and those most widely studied are included in Table 1. Work at
the biochemical and molecular level is probably more advanced than in
barley. Approximately 80 % of the zein polypeptides seperate into two
groups by SDS-PAGE, one having a molecular weight of 22,000 and the
other, 20,000 (Soave and Salamini, 1984). Sequence analysis of cDNA
clones specifying these two groups indicate that they are distantly related
to each other and that their structural genes probably originated from the
duplication and subsequent divergence of a single, ancestral gene (Marks
and Larkins, 1982).
The maize mutants lised in Table 1 show a wide range of zein inhi-
bition. Furthermore, specificity of inhibition can differ. Thus, opaque-7
primarily suppresses the 22,000 zein group, opaque-2 the 20,000 group,
whereas opaque-6 and floury-2 appear to suppress each protein group to
the same extent (Soave and Salamini, 1984). The reduced production of
zein in all the mutants shown in Table 1 is due to a reduced level of zein
messenger RNA in the endosperm cells (Langridge et al., 1982).
Soave et al. (1981) speculated that the loci regulating zein synthesis
might operate by producing factors which interfere with zein production
and, being diffusable, interact with all the zein genes dispersed in the maize
genome. A search was made for regulatory proteins using specially con-
structed lines which had various regulatory genes inserted in the same
genetic background as the selected low-lysine (control) line. It was demon-
strated that the control contained a major salt-soluble protein which was
virtually absent in the mutants. The authors provided evidence that the
protein was coded for at the opaque-6 locus and that it probably also
interacts in some unknown way with the opaque-2 locus.
In related experiments, Galante et al. (1983) showed that an albu-
min-type protein of molecular weight 70,000, present in small amounts in

controls, is overproduced in the floury-2 mutants. The protein occurs
mainly in protein bodies with some evidence that it is located on the
protein-body membrane. The authors suggested that the protein may
interfere with the transport and accumulation of zein in the protein bodies.
These results are of great interest but more work is required to elucidate
the precise mechanisms of action of these regulatory genes.
In wheat in particular, manipulation of both gene and chromosome
mutants will help in the understanding of the structure of glutenin and the
molecular basis for its elasticity. The HMW subunits of glutenin are
probably the main determinants of the elastic properties of dough (Payne
et ai., 1981; Tatham et ai., 1985). These subunits, as well as showing
extensive allelic variation (Fig. 3), are also occasionally deleted. Thus the
line shown in Fig. 7, slot 3, lacks a protein presumably because of a null
allele at Giu-l on chromosome 1 D and that in slot 9 a protein at the equiv-
alent locus on chromosome 1 B. When these two lines are crossed, a quarter
of the progeny will have double nulls. In the author's laboratory, a line is
..11.# __ __ ..
1 3 9
Fig. 7. SDS-PAGE of wheat landraces from Afghanistan. Deleted HMW subunits
of glutenin occur in slot 3 (chromosome 1 D) and slot 9 (chromosome 1 B). The
expected positions are arrowed
222 P. I. Payne
being developed by repetitive backcrossing which will have the same

glutenin subunit composition as the backcrossed parent except for the
deletion of four of the five HMW subunits. Similarly another line is being
produced, exploiting chromosome mutants with deleted Gli-l loci (Fig. 6)
and thus deficient in low-molecular-weight (LMW) glutenin subunits. Such
lines, with drastically altered proportions of glutenin subunit types, should
help in the elucidation of glutenin structure. Very recently, Graveland et al.
(1985) described a new model for glutenin structure in which subunit 10, a
chromosome 1 D-encoded HMW subunit, plays the key role in the
complex. Currently in the author's laboratory, a line is being developed
which lacks subunit 10, or any allelic variant of it, which otherwise has the
same subunit composition as a control line containing subunit 10. Analysis
of the two lines should enable the hypothesis of Graveland et al. (1985) to
be tested.
D. Role in the Food Industry

In the U.K. and many other countries, varieties of wheat and varieties of
barley have different end uses and it is essential that they do not become
intermixed prior to processing. Thus if bread-quality wheats which are
hard milling, strong mixing and have a high grain protein content were
significantly contaminated with biscuit quality wheats (soft milling, weak
mixing, low in protein) or feed wheats (hard milling, weak mixing, low in
protein) they would seriously affect the quality of loaves produced at the
bakery. Similarly maltsters and brewers must be able to distinguish malting
barleys from feed barleys.
For wheat there is not sufficient variation in the morphological char-
acters of grain to distinguish varieties. At the mill, rapid chemical tests can
be used to distinguish, for instance, hard-milling wheats from soft-milling
wheats and wheats rich and poor in protein, but for unequivocal identifi-
cation of varieties, protein electrophoresis is used. The standard, interna-
tional method used is electrophoresis of gliadins at pH 3.1-3.2 using
either starch or polyacrylamide gels (Ellis, 1984). Gliadins are coded by
genes at six, unlinked complex loci in the wheat genome and through spon-
taneous mutation there are many alleles at each locus (Payne et al., 1984 b).
The great majority of varieties can be distinguished by this technique and
those which cannot may be identified by an additional electrophoretic test,
SDS-PAGE (Shewry et al., 1978), which reveals allelic variation at three
more loci coding for HMW subunits of glutenin (Payne et aI., 1984b).
Thus, individual analyses on 100 grains can determine the varietal compo-
sition of an incoming shipment destined for bread-making. An unac-
ceptably high contamination by feed or biscuit-quality wheat would result
in its rejection at the mill (Ellis, 1984). In many Western European coun-
tries, North America and Australia formal descriptions of newly accepted
varieties include gliadin banding patterns, with varietal identification in
mind.
Unfortunately, identification of barley varieties by electrophoresis is
much less effective. For instance, Cooke et ai. (1983) were only able to
divide 68 barley cultivars into 19 groups on the basis of hordein compo-
sition. There are two main reasons for this: first, as barley is a diploid it
would be expected to contain only about one third the variation of that
found in wheat. Second, the two hordein gene loci are closely linked
(Shewry et ai., 1984a), thus making it likely that the final, elite line selected
from a varietal cross would have both B- and C-hordeins from one of the
parents. For these reasons, electrophoresis has not become nearly as
important to the maltster as it has to the miller.
VI. Chromosome Mutations
A chromosome mutation is usually defined as a structural change in the

genome resulting in the gain, loss or translocation of a chromosome
segment. Many types of such mutations, notably chromosome loss, can
only be tolerated in polyploid organisms so it is not surprising that the
most widely studied cereal is hexaploid wheat.
In wheat, deletions are a common form of chromosome mutation. Riley
and Kimber (1961) examined the mitotic chromosomes of root-tip nuclei
from 2027 seedlings in four wheat varieties. Twenty-two were atypical: 14
were monosomics, three were trisomies and one was telocentric. Similarly
in the other polyploid cereal, oats, a low level of abnormal karyotypes
occurs. These whole chromosome effects are most likely due to a failure of
chromosome pairing at meiosis.
Using the wheat variety Chinese Spring, Sears (1954) obtained a com-
plete series of lines monosomic for the 21 different chromosomes
(2 n = 41). These were either obtained as a result of spontaneous mutations
or from Chinese Spring lines nullisomic for chromosome 3 B, which
induces pairing failure in other chromosomes at meiosis.
Monosomic series have been used to determine the chromosomal
location of storage-protein genes in wheat (Payne et ai., 1980) but not
extensively because separate electrophoretic analyses of several grains are
required and conclusions are based partially on the relative amount of
protein under study which is produced in the developing grain. The reason
for this is that selfed grains of monosomics (those used for analysis) have
different chromosome numbers. On average, 73 % of the progeny are
monosomics, 24 % are disomics and 3 % nullisomics (Law and Worland,
1973). Thus, a protein not controlled by the monosomic chromosome will
be present at equal intensity in all the progeny analysed but proteins which
are controlled by genes on this chromosome will occur at 1, 3 and 0 dose
levels in the respective progeny (Payne et ai., 1980). Monosomic series are
now available for over 60 different varieties (Law et ai., 1983) so enabling
the chromosomal location of many protein variants to be determined.
Having established the complete series of Chinese Spring monosomics,
Sears (1954) went on to obtain, by selfing and selection, the complete series
of nullisomics. A complete series of trisomics (2 n = 43) was also obtained
224 P. I. Payne
by the same methods as those used to obtain monosomics. By selfing the

trisomics, tetrasomics (2 n = 44) were obtained and collected in the segre-
gating progeny.
None of these stocks have been used to any extent to study the genetics
of the storage proteins in wheat. Although nullisomic grains would be ideal
for locating genes to chromosomes they nearly all produce feeble plants
with few grains and several are sterile. Fortunately, Sears (1966) developed
a series of compensating nullisomic-tetrasomics (2 n = 42). In these lines a
chromosome pair (e. g. 1 A) is missing but the normal, euploid chro-
mosome number of 42 is maintained by the presence of an extra pair of ho-
moeologous chromosomes (either 1 B or 1 D). These plants have, in the
main, near-normal fertility and have been used extensively in locating
storage protein genes to chromosomes (for example, Wrigley and
Shepherd, 1973; Bietz et al., 1975; Brown et al., 1979).
As well as whole chromosome deletions or additions, Sears (1954)
developed various types of line for Chinese Spring with chromosome arms
deleted. The ditelocentrics, lacking either the short or long arms of a
chromosome pair, have been studied extensively to locate storage-protein
genes to chromosome arms.
Compensated nullisomic-tetrasomics and ditelocentrics can also be
used to advantage in recombination and linkage studies on storage protein
genes (see also Section V.B) but they have only been exploited rarely. In
one example, Payne et al. (1982) estimated recombination between storage-
protein loci Glu-Al and Gli-Al on the long and short arms respectively of
chromosome 1 A by analysing progeny of the cross:
Chinese Spring x Chinese Spring (Hope 1 A) (Primary cross)
~
FI 9 x Chinese Spring ditelo 1 AL d (Secondary cross)
In the primary cross, recombination occurs between the 1 A chromo-

somes of Chinese Spring and Hope. The FI plants were then crossed to a
ditelocentric line lacking the short arm of chromosome 1 A and therefore
the Gli-Allocus. Segregation of the chromosome 1 A encoded gliadins can
therefore be determined unambiguously by gel electrophoresis without
interference from the equivalent proteins from the male parent of the sec-
ondary cross, or from other proteins coded by genes on other chromo-
somes.
When ditelocentrics are used in the primary cross, the distance between
the centromere and any gene on the remaining arm can be estimated in
terms of recombination units, provided there is allelic variation. This type
of cross was used by Payne et al. (1982) to measure the distances from
Glu-Al, Glu-Bl and Glu-Dl to the centromere of the group 1 chromosomes
(see Fig. 6) and by Rybalka and Sozinov (1979) to map Gli-Bl to the cen-
tromere of chromosome 1B (see Fig. 6).
Chromosome deficiencies resulting from breakage during meiosis have
been described for chromosomes 6BL (Giorgi, 1981) and 5AL (Miller and
Reader, 1982). In the author's laboratory, chromosome deficiencies of the

homoeo10gous group 1 have been detected in about 10 out of several
thousand grains and two of them have been well characterised (Payne et
al., 1984a; Ainsworth et al., 1984). Chromosome mutants involving the loss
of storage-protein genes are easily detected by gel electrophoresis of
progeny from crosses. An example in which a segment of the short arm of
chromosome 1B containing Gli-BI has been deleted is shown in Fig. 1 of
Payne et al. (1984a). Lines with chromosome deficiencies have been very
useful in deletion mapping of various genes and in relating gene recombi-
nation maps with physical maps of chromosomes (Payne et al., 1984 a;
Snape et al., 1985).
An inter-chromosomal translocation has probably contributed to the
distribution of storage-protein genes in the wheat genome. Recent
sequencing of cloned genes for a-gliadins has shown a low but significant
sequence homology to y-gliadin genes (Rafalski et al., 1984; Kasarda et al.,
1984) and it is now generally recognised that they are derived from a
common ancestral gene. By comparison with other cereals (Shewry et al.,
1984a) it is clear that a chromosome homoeologous to the group 1 chromo-
somes of wheat carried the ancestral gliadin genes. It has been speculated
that the a-/~-gliadin genes were transferred from chromosome 1 to chrom-
osome 6 in wild diploid species by an interchromosomal translocation well
before the origin of hexaploid wheat (Shepherd and Jennings, 1970;
Kasarda, 1982). Interchromosomal translocations were probably also
responsible for the presence of secalin genes on chromosome 2R of rye,
Secale cereale (Shewry et aI., 1984 b) and 5R of S. montanum (Shewry et al.,
1985).
A chromosome inversion could account for the separation of hordein
genes on chromosome 5 of barley. They occur at two loci, about eight rec-
ombination units apart (Shewry et aI., 1984a), Hor 1 coding for C-hordein
and Hor 2 for B-hordein. Recent cDNA sequencing studies of these two
hordein groups (Forde et al., 1985) strongly suggest a common evolu-
tionary origin and so a common gene locus.
Oats are also hexaploid, containing three different genomes each with
seven pairs of chromosomes. Progress in assembling chromosome mutants
has been slow. By 1974, all 21 monosomics had been produced but in six
different genotypes (Rajhathy and Thomas, 1974). A disadvantage with
oats in the development of further genetic stocks is that selfing mono-
somics produces a variable number of nullisomics and for certain chromo-
somes none are produced. So far, no related work on the genetics of oat
globulin and prolamin has been fully reported although some work is now
in progress.
Some types of chromosomal mutants occur in diploid cereals although,
of course, lines with whole or partial chromosome deletions are not
available. For barley (2 n = 14) several complete sets of trisomics are
available for varieties Betzes (Eslick and Ramage, 1969) and Shin Ebisu
No. 16 (Tsuchiya, 1967). Monotelotrisomics are stable in several cereals
although complete sets have not yet been produced (Singh and Tsuchiya,
1977).
226 P. I. Payne
For maize, trisomics and interchromosomal translocation were ana-

lysed by electrophoresis amd it was shown that the zein genes are located
on chromosomes 4, 7 and 10 (Valentini et al., 1979). For barley, equivalent
stocks were used indirectly to determine the chromosome location of the
hordein genes. By recombination mapping, the hordein genes were shown
to be fairly closely linked to Ml-a genes giving resistance to powdery
mildew (Oram et al., 1975). Previously Ml-a had been shown using chro-
mosome mutants to be located on chromosome 5 (Burnham and Hagberg,
1956) which must also be the chromosome carrying the hordein genes.
VII. Conclusions
The storage-protein genes of cereals are tissue-specific in their output and

are developmentally regulated. That is they are derepressed only in the
endosperm (or for some genes additionally the embryo) and at only a spe-
cific period in the growth and development of this tissue. For these and
other reasons the genes have attracted the attention of molecular biologists
and significant advances have already been made in terms of the structure
and evolutionary relationships. The formidable tasks ahead in under-
standing the tissue specificity and regulation of these genes will very likely
be aided by the geneticist who, by using natural and induced variation in
the structural and regulatory genes, will be able to construct genotypes to
rigorously test future hypotheses on the molecular mechanisms of plant
development.
In wheat, a greater understanding of the relationship between bread-
making quality and allelic variation of storage-protein genes is needed and
this will be achieved from the analysis of specifically constructed genetic
lines. Molecular biologists are already close to understanding at the
sequence level why certain HMW subunits of glutenin impart greater elas-
ticity than their allelic counterparts (Flavell et al., 1984). In future, this
knowledge will be exploited in screening for good-quality glutenin sub-
units from diverse sources including landraces of ancient agriculture and
wild diploid species related to wheat. The genes coding for them would
then be transferred to commercial varieties by recurrent backcrossing.
Acknowledgements
I am extremely grateful to several of my colleagues at the Plant Breeding

Institute: Mrs. L. M. Holt for Fig. 1, Dr. E. A. Jackson for Fig. 2, Mr. D. B.
Smith for Fig. 3 and Dr. R. B. Austin for critically reading the text. Thanks
are also due to Dr. M. L. Parker (Food Research Institute, Norwich) for
permission to quote some of her unpublished work.
VIII. References
Ainsworth, C. c., Johnson, H. M., Jackson, E. A., Miller, T. E., Gale, M. D., 1984:
The chromosomal locations of leaf peroxidase genes in hexaploid wheat, rye
and barley. Theor. Appl. Genet. 69, 205-210.
Altschul, AM., 1965: Proteins: their chemistry and politics. London: Chapman
and Hall.
Bartels, D., Thompson, R. D., 1983: The characterisation of cDNA clones coding
for wheat storage proteins. Nuc. Acids Res. 11,2961-2977.
Bietz, J. A., Shepherd, K. W., Wall, J. S., 1975: Single-kernel analysis of glutenin:
used in wheat genetics and breeding. Cereal Chern. 52, 513-532.
Brown, J. W. S., Kemble, R J., Law, C. N., Flavell, R. B., 1979: Control of endo-
sperm proteins in Triticum aestivum (var. Chinese Spring) and Aegi/ops umbel-
lulata by homoeologous group 1 chromosomes. Genetics 93, 189-200.
Burnham, C. R., Hagberg, A, 1956: Cytogenetic notes on chromosome inter-
changes in barley. Hereditas 42, 467 -482.
Burr, F. A., Burr, B., 1981: In vitro uptake and processing of prezein and other
maize preproteins by maize membranes. J. Cell BioI. 90, 427 -434.
Cameron-Mills, v., Ingversen, J., Brandt, A, 1978: Transfer of in vitro synthesised
barley endosperm proteins into the lumen of the endoplasmic reticulum.
Carlsberg Res. Commun. 43, 91-102.
Cameron-Mills, V., von Wettstein, D., 1980: Protein body formation in the devel-
oping barley endosperm. Carlsberg Res. Commun. 45, 577-594.
Cooke, R J., Cliff, E. M., Draper, S. R, 1983: Barley cultivar characterisation by
electrophoresis II. Classification of hordein electrophoregrams. J. Nat. Inst.
Agric. Bot. (G. B.) 16, 197-206.
Doll, H., 1980: A nearly non-functional mutant allele of the storage protein locus
Hor2 in barley. Hereditas 93, 217-222.
Doll, H., 1983: Barley seed proteins and possibilities for their improvement. In:
Gottschalk, W., Muller, H. P. (eds.), Seed Proteins, pp. 207-223. The Hague:
Martinus Nijhoff.
Doll, R., 1984: Nutritional aspects of cereal proteins and approaches to overcome
their deficiencies. Philos. Trans. R. Soc. London, Ser.B 304, 373-380.
Ellis, J. W. S., 1984: The cereal grain trade in the United Kingdom: the problem of
cereal variety. Philos. Trans. R Soc. London, Ser.B 304, 395-407.
Eslick, R F., Ramage, R. T., 1969: Primary trisomies in the variety Betzes. Barley
Newsl. 12, 17.
Flavell, R B., Payne, P. I., Thompson, R. D., Law, C. N., 1984: Strategies for the
improvement of wheat-grain quality using molecular genetics. Biotechnol.
Genet. Eng. Rev. 2, 157-173.
Forde, B. G., Kries, M., Williamson, M. S., Fry, R. P., Pywell, J., Shewry, P. R.,
Bunce, N., Miflin, B. J., 1985: Short tandem repeats shared by B- and C-
hordein cDNAs suggest a common evolutionary origin for two groups of cereal
storage protein genes. EMBO J. 4, 9-15.
Galante, E., Vitale, A., Manzocchi, L., Soave, c., Salamini, F., 1983: Genetic
control of a membrane component and zein deposition in maize endosperm.
Mol. Gen. Genet. 192,316-321.
Giorgi, B., 1981: A line with a deletion on the long arm of chromosome 6B isolated
in Triticum aestivum cv. Chinese Spring. Wheat Inf. Servo 50, 22-23.
Grama, A, Gerechter-Amitai, Z. K., Blum, A., 1984: Breeding bread wheat cul-
228 P. I. Payne
tivars for high protein content by transfer of protein genes from Triticum dicoc-
coides. In: Cereal Grain Protein Improvement, pp. 145-153. Vienna: Interna-
tional Atomic Energy Agency.
Graveland, A., Bosveld, P., Lichtendonk, W. J., Marseille, J. P., Moonen, J. H. E.,
Scheepsta, A., 1985: A model for the molecular structure of the glutenins from
wheat flour. J. Cereal Sci. 3, 1-16.
Harlan, J. R., Starks, K. J., 1980: Germplasm resources and needs. In: Maxwell, F.
G., Jennings, R. (eds.): Breeding Plants Resistant to Insects, pp. 254-273. New
York: John Wiley.
Hu, N. T., Peifer, M. A., Heidecker, G., Messing, J., Rubenstein, I., 1982: Primary
structure of a genomic zein sequence of maize. EMBO J. 11, 1337-1342.
IAEA, 1984: Cereal Grain Protein Improvement, 385 pp. Vienna: International
Atomic Energy Agency.
Kasarda, D. D., 1982: Toxic proteins and peptides in celiac disease: relations to
cereal genetics. In: Walcher, D., Kretchmer, N. (eds.), Food, Nutrition and
Evolution. New York: Masson.
Kasarda, D. D., Okita, T. W., Bernadin, J. E., Baecker, P. A., Nimmo, C. C., Lew, E.
J.-L., Dietler, M. D., Greene, F. c., 1984: Nucleic acid (cDNA) and amino acid
sequences of a-type gliadins from wheat (Triticum aestivum). Proc. Nat. Acad.
Sci. U.S.A. 81, 4712-4716.
Koie, B., Doll, H., 1979: Protein and carbohydrate components in the Risq,
high-lysine barley mutants. In: Seed Improvement in Cereal and Grain
Legumes, I, pp. 205-215. Vienna: International Atomic Energy Agency.
Kries, M., Shewry, P. R., Forde, B. G., Rahman, S., Miflin, B. J., 1983: Molecular
analysis of a mutation conferring the high lysine phenotype on the grain of
barley (Hordeum vulgare). Cell 34, 161-167.
Kries, M., Shewry, P. R., Forde, B. G., Rahman, S., Bahramian, M. B., Miflin, B. J.,
1984: Molecular analysis of the effect of the lys 3a gene on the expression of
Hor loci in developing endosperms of barley (Hordeum vulgare). Biochem.
Genet. 22, 231-255.
Langridge, P., Pintor-Toro, J. A., Feix, G., 1982: Transcriptional effects of the
opaque-2 mutation of Zea mays L. Planta 156, 166-170.
Larkins, B. A., Hurkman, W. G., 1978: Synthesis and deposition of zein proteins in
maize endosperm. Plant Physioi. 62, 256-263.
Law, C. N., Worland, A. J., 1973: Aneuploidy in wheat and its use in genetic
analysis. In: Annual Report, 1972, pp. 25-65. Cambridge: Plant Breeding
Institute.
Law, C. N., Snape, J. W., Worland, A. J., 1983: Quantitative genetic studies in
wheat. In: Proc. 6th International Wheat Genetics Symposium, Kyoto, Japan
1983, pp. 539-547. Kyoto: Germplasm Institute.
Lawrence, G. J., Shepherd, K. W., 1981: Chromosomal location of genes con-
trolling seed proteins in species related to wheat. Theor. Appi. Genet. 59,
25-31.
Marks, M. D., Larkins, B. A., 1982: Analysis of sequence microheterogeneity
among zein messenger RNAs. J. BioI. Chern. 257,9976-9983.
Miflin, B. J., Shewry, P. R., 1979: The synthesis of proteins in normal and high
lysine barley seeds. In: Laidman, D. L., Wyn Jones, R. G. (eds.), Recent Adv-
ances in the Biochemistry of Cereals, pp. 239-273. London: Academic Press.
Miflin, B. J., Forde, B. G., Kreis, M., Rahman, S., Forde, J., Shewry, P. R., 1984:
Molecular biology of the grain storage proteins of the Triticeae. Philos. Trans.
R. Soc. London, Ser.B 304, 333-339.
Miller, T. E., Reader, S. M., 1982: A major deletion of part of chromosome 5A of

Triticum aestivum. Wheat Inf. Servo 55, 10-12.
Oram, R. N., Doll, H., Koie, B., 1975: Genetics of two storage protein variants in
barley. Hereditas 80, 53-58.
Orth, R. A., Bushuk, W., 1972: A comparative study of the protein of wheats of
diverse baking qualities. Cereal Chern. 49, 268-275.
Osborne, T. B., 1907: The proteins of the wheat kernel. Carnegie Inst. Washington
Publ. 84. Washington: Judd and Dutweiler.
Park, W. M., Stegemann, H., 1979: Rice protein patterns. Comparison by various
PAGE - techniques in slabs. Z. Acker Pflanzenbau 148,446-454.
Parker, M. L., 1980: Protein body inclusions in developing wheat endosperm. Ann.
Bot. 46, 29-36.
Parker, M. L., 1982: Protein accumulation in developing endosperm of a high-
protein line of Triticum dicoccoides. Plant Cell Environ. 5, 37 -43.
Payne, P.I., Corfield, K. G., 1979: Subunit composition of wheat glutenin proteins,
isolated by gel filtration in a dissociating medium. Planta 145, 83-88.
Payne, P. I., Law, C. N., Mudd, E. E., 1980: Control by homoeologous group 1
chromosomes of the high-molecular-weight subunits of glutenin, a major
protein of wheat endosperm. Theor. Appl. Genet. 58, 113-120.
Payne, P. I., Corfield, K. G., Holt, L. M., Blackman, J. A., 1981: Correlations
between the inheritance of certain high-molecular-weight subunits of glutenin
and bread-making quality in progenies of six crosses of bread wheat. J. Sci.
Food. Agric. 32, 51-60.
Payne, P. I., Rhodes, A. P., 1982: Cereal storage proteins: structure and role in agri-
culture and food technology. Encyl. Plant. Physiol. 14A, 346-369.
Payne, P. I., Holt, L. M., Worland, A. J., Law, C. N., 1982: Structural and genetical
studies on the high-molecular-weight subunits of wheat glutenin. Part 3. Telo-
centric mapping of the subunit genes on the long arms of the homoeologous
group 1 chromosomes. Theor. Appl. Genet. 63, 129-138.
Payne, P. I., Holt, L. M., Hutchinson, J., Bennett, M. D., 1984a: Development and
characterisation of a line of bread wheat, Triticum aestivum, which lacks the
short-arm satellite of chromosome IB and the Gli-Bllocus. Theor. Appl. Genet.
68,327-334.
Payne, P. I., Holt, L. M., Jackson, E. A., Law, C. N., 1984b: Wheat storage pro-
teins: their genetics and their potential for manipulation by plant breeding.
Philos. Trans. R. Soc. London, Ser.B 304, 359-371.
Peterson, D. M., 1978: Subunit structure and composition of oat seed globulin.
Rafalski, J. A., Scheets, K., Metzler, M., Peterson, D. M., Hedgcoth, C., SolI, D. G.,
1984: Developmentally regulated plant genes: the nucleotide sequence of a
wheat gliadin clone. EMBO J. 3, 1409-1415.
Rajhathy, T., Thomas, H., 1974: Cytogenetics of oats (Avena L). Ottawa: The Gen-
etics Society of Canada.
Rhighetti, P. G., Gianazza, E., Viotti, A., Soave, c., 1977: Heterogeneity of storage
proteins in maize. Planta 136, 115-123.
Riley, R., Kimber, G., 1961: Aneuploids and the cytogenetic structure of wheat
varietal populations. Heredity 16, 275-290.
Robert, L. S., Matlashewski, G. J., Adeli, K., Nozzolillo, C., Altosaar, I., 1983 a:
Electrophoretic and developmental characterisation of oat (Avena sativa L)
globulins in cultivars of different protein content. Cereal Chern. 60, 231- 234.
Robert, L. S., N ozzolillo, D., Altosaar, I., 1983 b: Molecular weight and charge
230 P. I. Payne
heterogeneity of prolamins (avenins) from nine oat (Avena sativa L) cultivars of

different protein content and from developing seeds. Cereal Chern. 60,
438-442.
Rybalka, A. I., Sozinov, A. A., 1979: Mapping the locus of Gli IB, which controls
the biosynthesis of reserve proteins in soft wheat. Tsitol. Genet. 13,276-282.
Sears, E. R., 1954: The aneuploids of common wheat. Research Bulletin Agricul-
tural Experimental Station, University of Missouri, Columbia, Missouri, 572.
Sears, E. R., 1966: Nullisomic-tetrasomic combinations in hexaploid wheat. In:
Riley, R., Lewis, K. R. (eds.), Chromosome Manipulation and Plant Genetics,
pp. 29-45. Edinburgh: Oliver and Boyd.
Shepherd, K. W., Jennings, A. c., 1970: Genetic control of rye endosperm proteins.
Experientia 27, 98-99.
Shewry, P. R., Faulks, A. J., Pratt, H. M., Miflin, B. J., 1978: The varietal identifi-
cation of single seeds of wheat by sodium dodecyl sulphate polyacrylamide gel
electrophoresis of gliadin. J. Sci. Food. Agric. 29, 847-849.
Shewry, P. R., Autran, J.-c., Nimmo, C. C., Ellen, J.-L., Kasarda, D. D., 1980: N-
terminal amino acid sequence homology of protein components from barley
and a diploid wheat. Nature (London) 286,520-522.
Shewry, P. R., Miflin, B. J., Kasarda, D. D., 1984a: The structural and evolutionary
relationships of the prolamin storage proteins of barley, rye and wheat. Philos.
Trans. R. Soc. London, Ser.B 304, 297-308.
Shewry, P. R., Bradberry, D., Franklin, J., White, R. P., 1984b: The chromosomal
locations and linkage relationships of the structural genes for the prolamin
storage protein (secalins) of rye. Theor. Appl. Genet. 69, 63-69.
Shewry, P. R., Miflin, B. J., 1985: Seed storage proteins of economically important
cereals. Adv. Cereal. Sci. Technol. 7, 1-83.
Shewry, P. R., Parmar, S., Miller, T. E., 1985: Chromosomal location of the struc-
tural genes for the Mr 75,000 y-secalins in Secale montanum Guss: evidence for
a translocation involving chromosomes 2R and 6R of cultivated rye (Secale
cereale). Heredity 54, 381-383.
Simmonds, D. H., O'Brien, T. P., 1981: Morphological and biochemical devel-
opment of the wheat endosperm. Adv. Cereal Sci. Technol. 4, 5-70.
Singh, N. K., Shepherd, K. W., 1985: The structure and genetic control of a new
class of disulphide-linked proteins in wheat endosperm. Theor. Appl. Genet.,
71,79-92.
Singh, R. J., Tsuchiya, T., 1977: Morphology, fertility and transmission in seven
monotelotrisomics of barley. Z. Pflanzenzucht. 78, 327-340.
Snape, J. W., Flavell, R. B., O'Dell, M., Hughes, W. G., Payne, P. I., 1985: Intra-
chromosomal mapping of the nucleolar organiser region relative to three
marker loci on chromosome 1B of wheat (Triticum aestivum). Theor. Appl.
Genet. 69, 263-270.
Soave, c., Tardani, L., Di Fonzo, N., Salamini, F., 1981: Zein level in maize endo-
sperm depends on a protein under control of the opaque-2 and opaque-6 loci.
Cell 27, 403-410.
Soave, c., Sa1amini, F., 1984: Organisation and regulation of zein genes in maize
endosperm. Philos. Trans. R. Soc. London, Ser.B 304, 341-347.
Tallberg, A., 1984: Biochemical and genetic characterisation of lysine genes and
their utilization in breeding barley for improved grain protein. In: Cereal Grain
Protein Improvement, pp. 205-214. Vienna: International Atomic Energy
Agency.
Tatham, A. S., Miflin, B. J., Shewry, P. R., 1985: The ~-turn conformation in wheat
gluten proteins: relationship to gluten elasticity. Cereal Chern., 62, 405-411.
Thompson, R. D., Bartels, D., Harberd, N. P., Flavell, R. B., 1983: Characterisation
of the multigene family coding for HMW glutenin subunits in wheat using
cDNA clones. Theor. Appi. Genet. 67, 87-96.
Tsuchiya, T., 1967: The establishment of a trisomic series in a two-rowed cultivated
variety of barley. Can. J. Genet. Cytoi. 9, 667 -682.
Valentini, G., Soave, c., Ottaviano, E., 1979: Chromosomal location of zein genes
in Zea mays. Heredity 42,33-40.
Vasal, S. K., Villegas, E., Tang, C. Y., 1984: Recent advances in the development of
quality protein maize germplasm at CIMMYT. In: Cereal Grain Protein
Improvement, pp. 167-189. Vienna: International Atomic Energy Agency.
Vogel, K. P., Johnson, V. A., Mattern, P. J., 1983: Results of systematic analyses for
protein and lysine composition of common wheats (Triticum aestivum L) in the
USDA world collection. Nebr. Res. Bull. 258, 27.
Wall, J. S., 1979: The role of wheat proteins in determining baking quality. In:
Laidman, D. L., Wyn Jones, R. G. (eds.), Recent advances in the Biochemistry
of Cereals, pp. 275 - 311. London: Academic Press.
Wrigley, C. W., Shepherd, K. W., 1973: Electrofocussing of grain proteins from
wheat genotypes. Ann. N. Y. Acad. Sci. 209, 154-162.
Yamagata, H., Tanaka, K., Kasai, Z., 1982: Evidence for a precursor form of rice
glutelin subunits. Agric. BioI. Chern. 46, 321-322.
Zhae, W. M., Gatehouse, J. A., Boulter, D., 1983: The purification and partial
amino acid sequence of a polypeptide from the glutelin fraction of rice grains:
homology to pea legumins. FEBS Lett. 162, 96-102.
Chapter 8
Molecular Approaches to Plant and Pathogen Genes
Richard I. S. Brettell and Anthony J. Pryor
CSIRO, Division of Plant Industry, P.O. Box 1600, Canberra City,

ACT 2601, Australia
Contents
I. Introduction
II. A Molecular Approach to Gene-for-Gene Resistance
A. The Shotgun Method
B. Transposon Mutagenesis or Gene Tagging
III. The Role of Toxins in Plant Disease
IV. Conclusions
V. References
I. Introduction
Host-pathogen combinations among higher plants cover a continuous wide

range of diversity. At one end is the extreme specialisation of obligate bio-
trophs such as the rust and mildew fungi, whose existence depends on an
association with living tissue of the host. At the other end are the facul-
tative necrotrophs which are able to grow in the absence of host tissue but
under certain circumstances will infect a plant and feed off the dead and
dying tissues.
These types of interaction have been well studied in descriptive terms,
in genetics, and to a lesser extent in biochemical terms, but have received
little attention from the recent development and application of molecular
biological techniques. In this article we will discuss some areas of plant-
pathogen interaction that are currently undergoing analysis by molecular
techniques and attempt to point to other promising approaches.
For the obligate biotrophs, many examples of a gene-for-gene rela-
tionship have been established since the studies by Flor (1947, 1956) on
flax and the rust Melampsora lini. Resistance in the host is conditioned by a
single, usually dominant, gene which matches a gene for avirulence in the
pathogen. There is a wealth of genetic information about major gene resis-
tances to a number of obligate biotrophs, yet an almost complete ignorance
234 R. I. S. Brettell and A. 1. Pryor
of the nature of the products of these genes and how they interact with the
genes of the pathogens. It is frequently noted that gene-for-gene resistance
is not the only level of plant resistance that is important. This is undoubt-
edly true and there is perhaps no better illustration of this than the demon-
stration by Wynn (1976) of the role of leaf epidermal topography on the
ability of a rust germ tube to locate a stomate and infect the host plant.
There must be a variety of levels, of varying importance, which all con-
tribute to effective resistance. However, molecular analysis is limited at
present to phenotypes with single gene inheritance, and traits such as leaf
surface topography are unlikely to fall into this class.
Moving from biotrophic to necrotrophic diseases, it is apparent, in one
sense, that the necrotrophic pathogens are less specialised than the obli-
gately parasitic organisms, in not needing to maintain the host cells in a
metabolically active condition. However, given the limited ranges of most
necrotrophs, virulence to a particular host plant must still be considered
the exception rather than the rule and, therefore, represents a specialised
level of adaptation. This in turn is consistent with genes in the host that
condition resistance and genes in the pathogen that condition virulence
being critical factors in disease development, even though the result is
scarcely a balanced relationship. In some cases virulence may be ascribed
to toxic telepathogenic substances produced by the pathogen, and such
examples may provide useful models for the study of host-pathogen gene
interactions.
At the molecular level, perhaps the best characterised plant and
pathogen relationship involves the bacterium Agrobacterium tumefaciens.
This is the causal agent of crown gall disease in a number of dicotyle-
donous plants, which results from infection by the bacterium at a wound
site and transfer of a small piece of DNA, the T-DNA, from a plasmid
within the bacterium to the host cell (Gheysen et at., 1985). This T-DNA is
integrated into the host cell genome and signals the proliferation of host
tissue through the production of phytohormones. Recent work has esta-
blished that the T-DNA carries genes that code for enzymes which are
involved in auxin and cytokinin biosynthesis and which represent new syn-
thetic pathways which cannot be regulated by the host (Schroder et at.,
1984; Inze et at., 1984; Buchmann et at., 1985). Other genes on the T-DNA
alter the host cell's metabolism by coding for enzymes involved in the syn-
thesis and catabolism of opines, a class of arginine derivatives which can
be used as an energy source by the bacterium. This may be a unique
example of parasitism at the DNA level, as there is no evidence that
transfer of genetic material from pathogen to host is a widespread pheno-
menon. Nevertheless, it serves to illustrate that the relationship between
host and pathogen depends essentially on an interaction between host
DNA and pathogen DNA; the intermediates in this interaction may be
RNA, proteins or substances produced as a result of the activity of specific
proteins, yet the balance between the two organisms rests at the DNA level.
Molecular Approaches to Plant and Pathogen Genes 235
II. A Molecular Approach to Gene-for-Gene Resistance
The gene-for-gene hypothesis, first elaborated in the analysis of the inher-

itance of resistance and virulence respectively in flax and its rust
Melampsora lin; (Flor, 1956), is the model system most readily accessible to
presently available molecular techniques. Perhaps the most glaring
anomaly in our understanding of plant disease is the discrepancy between
our knowledge of the genetics and the almost complete absence of infor-
mation of the underlying biochemical mechanisms. Most major gene
resistance to plant pathogens has its expression in the hypersensitive
response, yet the mechanisms that initiate this response are unknown
(Hooker, 1967; Ingram, 1978). Several attempts (Torp and Andersen, 1982;
Manners et al., 1985) to analyse the early events in the resistance response
all suffer from the difficulty involved in distinguishing cause from effect.
Molecular analysis of plant genes has focussed mainly on genes with
abundant or easily recognisable gene products. This approach has not been
possible for gene-for-gene resistance and methods must be found which do
not depend on a prior knowledge of the gene products involved.
As stated by Flor (1956), the gene-for-gene hypothesis concluded that
"for each gene conditioning rust reaction in the host, there is a specific
gene conditioning pathogenicity in the parasite". With few exceptions,
both the host resistance gene and the pathogen avirulence gene specifically
recognised in the incompatible or resistant reaction are dominant. Some of
the apparent exceptions are due to genetic complexities superimposed on
the basic gene-for-gene model. For example, dominant virulence in flax
rust is due to an independent locus which is a dominant inhibitor of the
avirulence gene function (Lawrence et al., 1981). However, for the vast
majority of gene-for-gene resistances the resistant reaction seems to occur
when the product of an avirulence gene is recognised by a product of an
allele of a host resistance gene. This view leads to several expectations
crucial in the development of methods of molecular analysis.
1. Transformation with a dominant gene should result in the acquisition
of a new function. Thus by transformation it should be possible to change
a susceptible plant to a resistant one and a virulent pathogen into an avir-
ulent race, but the reciprocal conversions are not expected.
2. Mutation of a dominant gene should result in a loss of function. Thus
a resistant plant will become susceptible and an avirulent pathogen will
become virulent.
At present there are two general methods that exploit these expecta-
tions. The first, called the "shotgun method" because of its inherently
random nature, involves random cloning of genomic DNA followed by
transformation and selection of the newly acquired character. The second
method is called "transposon mutagenesis" in micro-organisms or "gene
tagging" in higher plants. The success of both methods depends on rare
events and on the absolute size of the genome. It is not surprising then that
these approaches have been successfully employed in the molecular
analysis of avirulence genes of pathogens, which have small genomes, but
236 R. I. S. Brettell and A. J. Pryor
not in the cloning of a resistance gene from a host plant with its much
larger genome size.
A. The Shotgun Method

The general approach is to produce a genomic library of the pathogen or
the host plant DNA, to use this library of DNA sequences to transform
cells of the host or pathogen and then to screen for the acquisition of
resistance or avirulence respectively. At present this can be done in a
cosmid vector in E. coli. Such vectors are based on the ability of bacterio-
phage Lambda to package 37-53 kb of DNA, with about 10-25 kb of
DNA comprising the vector specific sequences and the remaining
20-40 kb being available for library sequences. Table 1 compares the
number of clones of transformants required to be 95 % certain of recov-
ering a gene for resistance from several plant species, together with yeast
and E. coli. Clearly from a consideration of genome size alone, the ability
to isolate avirulence genes from plant pathogens such as fungi or bacteria
is going to be a much simpler job than isolating resistance genes from plant
hosts with their much larger genomes.
Recently, Staskawicz et al. (1984) have used this approach to isolate a
bacterial avirulence gene which determines incompatibility on resistant
soybean leaves. The bacterium, Pseudomonas syringae pv. glycinea, is the
causal agent for bacterial blight of soybean, and host resistance is thought
to be of the gene-for-gene type already outlined. A cosmid library, with an
average insert size of about 25 kb, was constructed from restriction frag-
ments of the DNA isolated from an avirulent race of the bacterium. Stas-
kawicz et al. first demonstrated that the particular cosmid vector could be
mobilized to various races of P.s. pv. glycinea, permitting the cosmid library
to be transduced into a virulent bacterial race and DNA sequences
Table 1. Number of transformants needed to contain a resistance gene
Haploid Genome Size bNumber of Transformants (N)

Species' x 106 bp 40kb 25 kb
Wheat 17,300 1,296,000 2,073,000

Rye 8,300 622,000 995,000
Barley 5,500 412,000 659,000
Pea 4,800 360,000 552,000
Maize 3,900 292,000 467,000
Flax 700 52,000 84,000
Yeast 30 2,200 3,600
E. coli 7.5 560 890
Sources are from Bennett and Smith (1976) and Bennett et al. (1982), and on
the approximation that I pg of DNA is equivalent to about I x 109 bp.
b The number of transformants (N) needed to give a 95 % chance of recovering a
resistance gene is calculated according to Clark and Carbon (1976) for cosmid
inserts of two sizes, 25 and 40 kb.
screened for the avirulence phenotype. Bacterial colonies, each containing

a different hybrid cosmid derived from the virulent race, were inoculated
onto soybean leaves from a plant susceptible to the virulent race but
resistant to the avirulent race whose DNA was used for the construction of
the cosmid library. Provided the avirulence gene was expressed in the reci-
pient bacterial cells, the expectation was that the rare transduced colonies
would cause a resistant reaction on the leaf. This would be readily visible
as a hypersensitive reaction of plant cells in the region of the inoculation.
From a consideration of the above table it is evident that Staskawicz et al.
needed to screen around 500-1000 cosmid clones to have a chance of rec-
overing the avirulence gene. In fact a single clone was recovered from the
680 independently transduced bacterial colonies that were tested. This
27.2 kb clone was also mobilized into other bacterial races and conferred
on these transconjugants the same avirulence phenotype that was present
in the parental or donor race. Thus the shotgun method works for small
genomes for which methods of genetic manipulation such as transfor-
mation are available. Unfortunately for most of the obligate pathogens
such as rusts and smuts these techniques are not available.
When it comes to using the shotgun method for cloning a plant
resistance gene the problems are even greater, not only because of the large
number of cosmid clones that need to be screened, but also because of the
difficulty in producing a sufficient number of plant transformants. Trans-
formation of dicotyledons is now feasible particularly with methods based
on the Ti vector system (Gheysen et ai., 1985) and improvements are being
added continually. The initial fear that expression might be an insurmount-
able problem has not been realised since the available evidence suggests
that genes present in transformed DNA will be expressed if they are pro-
vided with appropriate 5' promoter and 3' terminal sequences (Herrera-
Estrella et al., 1984). Transformation of monocotyledons has not yet
reached this stage of development but the difficulties of transformation
with gene expression for angiosperms in general are likely to be overcome
in the near future. However, this still leaves the problem of the enormous
number of transformants that have to be screened to recover a unique gene
sequence. Either the size of the DNA in the (cosmid) library must be sub-
stantially increased or methods must be developed to fractionate or to
decrease the size of the genome. Working with small genomes such as that
of flax will help and this could be enhanced by using donor DNA from a
plant carrying more than one functional resistance gene. For example,
there are five loci that confer resistance to flax rust.
In plants with larger genomes it might be possible to fractionate whole
chromosomes if methods for isolating populations of metaphase chromo-
somes were developed. In maize the tip of the short arm of the smallest
chromosome (chromosome 10) carries a gene complex of four to five loci
that confer rust resistance (Hooker, 1967). Furthermore, the tip of this
chromosome, consisting of a small region of euchromatin containing the
genes for rust resistance, can be translocated to a B chromosome (Beckett,
1978). If this euchromatin could be separated from the heterochromatin in
a preparation of purified B chromosomes and used to create the cosmid

library, the number of transformants needed to recover a resistance gene
would be reduced 100 fold.
B. Transposon Mutagenesis or Gene Tagging

Transposon mutagenesis in bacteria results in inactivation of a gene by the
insertion of a transposable element into the gene. Generally the transposon
carries a selectable drug resistance marker. Thus if drug resistant mutations
of avirulence to virulence are selected in a pathogen, they are probably the
result of insertional inactivation of the avirulence gene. This mutation of
the avirulence gene is now accessible to molecular analysis without any
prior knowledge of the function of the gene. A genomic clone of the
mutation can be isolated on the basis of homology to the transposon base
sequences. Flanking sequences, those sequences adjacent to but not homol-
ogous to the transposon sequences, should be sequences normally involved
in the expression of the avirulence phenotype and can be used to isolate
the intact gene. Although trans po son mutagenesis can be considered part
of the normal repertoire for the genetic and molecular analysis of micro-
organisms, it has yet to see widespread application to the analysis of plant
pathogens. The technique has been used to isolate sequences essential for
pathogenesis in strains of Pseudomonas syringae (Anderson and Mills,
1985; Niepold et al., 1985) but these may have little relationship to aviru-
lence genes involved in pathogen specificity such as occurs in the gene-
to-gene system. Many gene products must be required for pathogen
growth. Bacteria of the genus Rhizobium might not be considered
pathogenic, however the use of transposon mutagenesis to isolate genes
required for nodulation is perhaps one of the best examples of the pot-
ential of this approach (see de Bruijn and Lupski [1984] for a general dis-
cussion of the use of transposon Tn 5 mutagenesis). The major difficulty in
the analysis of avirulence will be in transferring the genetic technology to
pathogenic organisms. As we have already seen from the preceeding dis-
cussion, this has been possible with a number of non-obligate pathogens,
particularly bacteria, but it is difficult to see how this technology will be
developed in an obligate biotrophic pathogen such as a plant rust.
In plants the equivalent approach to transposon mutagenesis is called
"gene tagging" and is at present only feasible in one plant species, Zea
mays. The first transposable genetic elements described by McClintock
were those in maize plants (see Federoff, 1983 for a recent review) and,
while similar elements certainly occur in other plants (Fincham and Sastry,
1974), only in maize has their analysis reached a stage where they can be
used as gene tags. The feasibility of this approach was demonstrated by the
isolation of the bronze 1 locus from maize (Federoff et al., 1984a) using an
inactive mutation of the gene which experimentally had been determined
to be a result of insertion of the transposable element Ac.
Taking this example as a model, it should be possible to isolate a
resistance gene from maize if certain further conditions are met. The
resistance must be specified by a single dominant gene and the system must
be amenable to mutagenesis by a transposable element system. The Rp 1
gene complex specifying resistance against Puccinia sorghi, the causal agent
of maize rust (Hooker, 1967), is a good example, and so too are resistances
to the He-toxin of Cochliobolus carbonum (Nelson and U11strup, 1964).
Next, it is necessary to recover a susceptible mutation of the resistance
gene and show that it is due to insertional mutagenesis by a given trans-
posable element. Unlike the bacterial transposons, the maize elements do
not carry a selectable drug-resistance marker and recovery of a mutation
requires the screening of a large number of progeny. Such a screen will
extract all susceptible mutations caused by a variety of mechanisms,
including those due to insertion of a transposable element. These may be
difficult or impossible to distinguish except that insertional mutations are
expected to be unstable (Freeling, 1984), reverting to the resistance pheno-
type because of the propensity of the element to undergo secondary trans-
position events. A number of independent transposable element systems
have been identified in maize (Federoff, 1983) and many may be useful as
gene tags, although, at present cloned probes are only available for a
limited few (Federoff et al., 1984b; Freeling, 1984; Pereira et al., 1985). At
present it is not clear whether all target loci in the plant are equally acces-
sible to insertional mutagenesis or even at what frequency such mutations
might occur. Unlike the internal portion of the Ac element, many trans-
posable elements are present in high copy number (> 50) (Freeling, 1984;
Sutton et al., 1984) and this can superimpose an additional difficulty in dis-
tinguishing these background genomic locations from genomic clones in
which the element is inserted into the resistance gene. To clone the allocus
from maize, O'Reilly et al. (1985) overCame this difficulty by isolating
clones from independent genomic libraries made from two insertional
mutations of the al locus which had been induced by the unrelated trans-
posable elements En and Mu 1. The En and Mu I selected clones were cross
hybridized and clones which had sequences in common could be used to
identify those clones carrying all or part of the al gene.
III. The Role of Toxins in Plant Disease
Most plants are resistant to most pathogens. Mechanisms of defence

include preformed barriers such as the cuticle and existing chemical sub-
stances which antagonise the growth of potential pathogens. Alternatively,
the barriers to disease development may be generated in response to the
presence of the pathogen and these induced barriers may also be structural
or chemical in nature. Specific chemical compounds produced by plants in
response to infection have been termed phytoalexins (Muller, 1956) and
have been identified in a range of plant species (Kuc, 1972). The onus is
then on the pathogen to circumvent the components of defence carried by
the host plant. The production of phytotoxic substances by a pathogen may
be seen as part of an overall strategy to get beyond host systems of defence,
240 R. I. S. Brettell and A. 1. Pryor
particularly those formed in response to infection. Not surprisingly the

production of phytotoxins is seen more often among necrotrophic
organisms where the host-pathogen interaction does not depend on the
plant cells remaining metabolically active.
For a small number of plant diseases, the pathogen is found to produce
a host-specific or host-selective toxin which is an absolute determinant of
pathogenicity_ Thus, production of the toxin is required for full disease
symptoms to develop on a susceptible host plant. Examples of pathogens
in this category have been collated by Yoder (1980), Kono et al. (1981),
Scheffer (1983) and Nishimura and Kohmoto (1983) and predominantly
comprise species of the fungi Alternaria and Drechslera (Cochliobolus). The
necessity of the toxin for disease development may be seen both for the
pathogen and for the host. Mutants of the pathogen which fail to produce
the host-specific toxin will either not infect the plant or at most produce
minor disease symptoms. This is illustrated by the experiments of Scheffer
et al. (1967) on Cochliobolus victoriae (perfect stage of Drechslera, or Hel-
minthosporium victoriae) and Cochliobolus carbonum (perfect stage of
Drechslera zeicola, Helminthosporium carbonum). C. victoriae produces the
host-specific HV-toxin and is the causal agent of Victoria blight of oats,
whereas C. carbonum race 1 produces HC-toxin and causes leaf spot of
maize. The fungi are heterothallic and sexually compatible, and progeny of
crosses were found to segregate in a manner consistent with the production
of toxin being under the control of a single dominant locus in each species.
In virtually all cases tested, pathogenicity of the resulting isolates was
correlated with the ability to produce toxin specific to the host.
Genotypes of the host which are resistant to the effects of toxin are also
resistant to the development of the particular disease. Susceptibility to Vic-
toria blight of oats is associated with a dominant gene governing resistance
to crown rust caused by Puccinia coronata (Welsh et al., 1954). The locus
for sensitivity to HV-toxin appears to be either allelic with or very tightly
linked to the rust resistance gene Pc-2 (see Day, 1974). In maize, resistance
to the HC-toxin of C. carbonum race 1 is determined by the two genes, Hm
on chromosome 1 and Hm-2 on chromosome 9 (Nelson and Ullstrup,
1964), whereas sensitivity to T-toxin of Cochliobulus heterostrophus (perfect
stage of Drechslera, or Helminthosporium victoria e) race T is associated with
a genetic alteration in the mitochrondrial genome (Leaver and Gray, 1982;
Laughnan and Gabay-Laughnan, 1983). In tomato, resistance to Alternaria
alternata f. sp. lycopersici is controlled by a single dominant nuclear gene
which also conditions resistance to AL-toxin(s) produced by the fungus
(Gilchrist and Grogan, 1976).
Many questions still remain concerning the relationship between the
pathogen and host genes. Among related pathogenic fungi, such as the
Cochliobolus species discussed above, there is as yet no clear structural
relationship between the host-specific toxins produced (Macko, 1983),
although the apparent chemical instability of HV-toxin has until now pre-
cluded a structural determination (Kono et al., 1981). It is thus not clear
whether the toxin production represents an amplification of an existing
pathway of secondary metabolism or a completely novel synthetic

pathway. In addition, there is no simple pattern regarding the sites of
action of the toxins in the respective host cells (Yoder, 1980; Daly, 1981).
Further genetic investigation of both pathogen and host should be able to
provide some of the answers, particularly as it becomes possible to
examine genetic differences at the DNA level. For example, in the Cochli-
obolus species where the production of toxin appears to be determined by
few genes, it would be informative to perform some molecular analysis of
those genes and their polypeptide products. Such genes might be identified
following transformation of avirulent toxin-less strains of a fungus with a
related toxin-producing strain. For instance, there is good evidence that a
single gene, toxl, determines virulence and T-toxin production in C. hetero-
strophus (Leach et al., 1982). Thus it should be possible to screen for toxin
production among homozygous toxl- lines genetically transformed with
cloned DNA from a toxl+ strain, and thereby identify at a structural level a
major determinant of toxin production. If successful, this approach might
provide evidence for or against the hypothesis that T-toxin production and
the resulting virulence of race T of C. heterostrophus is due to a recent
mutation in race 0 (Leonard, 1973), thus shedding light on the evolution of
the capacity to produce a host-specific toxin.
Similarly, in the host plant, application of a molecular approach
may make it possible to determine precisely the genetic difference
between resistant and susceptible lines. Steps have been made in this
direction for the southern com leaf blight disease caused by C. heter-
ostrophus, where susceptible T-toxin sensitive lines of maize invariably
carry the Texas (T) source of cytoplasmic male-sterility (Hooker et al.,
1970; Hilty and Josephson, 1971). Mitochondria from T cytoplasm
were found to be very sensitive to T-toxin of C. heterostrophus at con-
centrations which had no effect on mitochondria from N (normal) cyt-
oplasm (Miller and Koeppe, 1971). Restriction fragment differences
have been identified between DNA of T-toxin sensitive mitochondria
from T cytoplasm and mitochondrial DNA from T-toxin resistant
revertants derived from tissue culture, although not all of the differ-
ences correlated with the change in toxin sensitivity (Gengenbach et
al., 1981; Kemble et al., 1982). The polypeptide translation products of
isolated mitochondria have also been examined and found to have dif-
ferences associated with the toxin sensitive phenotype (Forde and
Leaver, 1980; Dixon et al., 1982). However, identification of the site
of action of T-toxin needs further information regarding the structure
and function of the alterations observed at the molecular level.
The link between HV-toxin sensitivity and crown rust resistance at the
Pc-21ocus may also be amenable to molecular analysis once techniques for
cloning single disease-resistance genes become generally available (see pre-
ceding section). Transformation studies using cloned fragments from and
adjacent to the Pc-2 locus should allow identification of the DNA
sequences responsible for imparting HV-toxin sensitivity to Victoria blight
susceptible oat plants.
In contrast to host-specific toxins, there are a number of necrotrophic

plant pathogens where the role of toxins is less clear. Here toxic substances
produced by the pathogen, although not absolute determinants of pathoge-
nicity, have been implicated as having an important function in disease
development (Scheffer, 1983). These include fungal pathogens such as Sep-
toria nodorum (perfect stage of Leptosphaeria nodorum), the causal agent of
glume blotch in wheat; Rhyncosporium secalis, the agent of barley scald
disease; Cochliobolus miyabeanus (perfect stage of Drechslera, or Helmin-
thosporium oryzae); and bacterial pathogens such as Pseudomonas syringae
pv. tabaci and Pseudomonas syringae pv. phaseolicola.
The fungal pathogen Septoria nodorum can be considered as an
example of how a genetical approach could help determine the contri-
bution of toxin production to disease development. There is evidence that
toxin substances produced by the fungus are to some extent determinants
of pathogenicity (Bousquet and Skajennikoff, 1974; Kent and Strobel,
1976). Tolerance to the disease in wheat has been correlated with resistance
of plant tissues to the effects of sterile culture filtrates of the fungus (Hann,
1978). Yet there are several difficulties associated with determination of the
importance of toxins in development of the disease. First, genetic analysis
is made difficult by the unavailability of wheat genotypes carrying full
resistance to the pathogen (Shipton et al., 1971; Bockmann et aI., 1975).
Second, disease development is influenced by factors such as architecture
and development of the plant. Tall and late-maturing genotypes tend to
escape infection (Bronnimann et al., 1973; Tavella, 1978; Scott et al., 1982).
Third, there is evidence for a degree of host specialisation among different
isolates of the fungus (Rufty et al., 1981). The absence of near isogenic host
material differing only in disease susceptibility is a serious disadvantage.
One approach to resolving this difficulty is to use toxins produced by the
fungus to select for resistance in cultured tissues (Brettell and Ingram,
1979), as was done in the selection of alfalfa with resistance to Fusarium
oxysporum (Hartman et al., 1984), another pathogen for which non-specific
toxins may have a role in disease development. A second possibility would
be to identify resistance in wheat lines carrying addition chromosomes
from related alien grass species. If such resistance was confined to a single
locus, then it would be feasible to examine how this resistance interacts
with pathogen genes or gene products. A truly integrated approach would
also involve genetic and biochemical analysis of the pathogen which, in
contrast to studies of biotrophic fungi such as the rusts and mildews, would
be made easier by axenic culture of the pathogen.
IV. Conclusions
We have reviewed just a few of the many possibilities for the application of
molecular techniques to the study of host-pathogen interactions. We are
not recommending a wholly molecular approach but rather trying to make
the obvious point that a number of relatively new molecular techniques may
provide fresh insights into what has been a largely intractable problem.
These techniques will help attack, but not necessarily solve, the problem.
Cloning a gene for toxin production will not tell us the function of this
gene without some more detailed understanding of the biochemistry and
the biology of the host-pathogen interaction. Recognizing differences in
restriction fragments between DNA of T-toxin sensitive mitochondria and
normal mitochondria provides a clear direction for future work to identify
the site of action of T-toxin.
Shotgun cloning and gene tagging are methods which could provide
isolated DNA sequences specifying resistance and avirulence gene func-
tions without any prior knowledge of the functions of these genes. The
determination of these functions will be the next step. The base sequences
may contain some obvious signals, such as an open reading frame indi-
cating that the gene product is a protein, something which is not known at
the moment. This sequence may code for certain protein structural features
which indicate a membrane location or a particular enzymatic function.
The gene sequence or a cDNA clone of the mRNA might be placed into an
expression vector producing sufficient gene product to elicit antibody pro-
duction which could be used for cytochemical localization studies. But
knowing that a gene product is, for example, a membrane located protein,
is only a part of the understanding of how a pathogen and its host interact.
Acknowledgement
The authors wish to thank their colleagues at CSIRO Division of Plant Indus-
try for their helpful comments during the preparation of the manuscript.
v. References
Anderson, D. M., Mills, D., 1985: The use of transposon mutagenesis in the iso-
lation of nutritional and virulence mutants in two pathovars of Pseudomonas
syringae. Phytopathology 75, 104-108.
Beckett, 1. B., 1978: B-A translocations in maize. I. Use in locating genes by
chromosome arms. 1. Hered. 69, 27-36.
Bennett, M. D., Smith, 1. B., 1976: Nuclear DNA amounts in angiosperms. Philos.
Trans. R. Soc. London, Ser.B 274, 227-274.
Bennett, M. D., Smith, 1. B., Heslop-Harrison, 1. S., 1982: Nuclear DNA amounts
in angiosperms. Proc. R. Soc. London, Ser.B 216, 179-199.
Bockmann, H., Mielke, H., Wachholz, G., 1975: Investigations on the susceptibility
of various winter and spring wheat cultivars to Septoria nodorum Berk. and
Fusarium culmorum Link. Z. Pflanzenzucht. 74, 39-47.
Bousquet, 1. F., Skajennikoff, M., 1984: Isolement et mode d'action d'une phyto-
toxine produite en culture par Septoria nodorum Berk. Phytopathoi. Z. 80,
355-360.
Brettell, R. I. S., Ingram, D. S., 1979: Tissue culture in the production of novel
disease-resistant crop plants. BioI. Rev. 54, 329-345.
Bronnimann, A., Fossati, A., Hani, E, 1983: Spreading of Septoria nodorum Berk.
and damage to artificially induced straw-length mutants of the winter wheat
Zenith (Triticum aestivum L.). Z. Pflanzenzucht. 70, 230-245.
244 R. I. S. Brettell and A ..J. Pryor
Buchmann, I., Marner, F. J., Schroder, G., Waffenschmidt, S., Schroder, J., 1985:
Tumour genes in plants: T-DNA encoded cytokinin biosynthesis. EMBO J. 4,
853-859.
Clarke, L., Carbon, J., 1976: A colony bank containing synthetic Col El hybrid
plasmids representative of the entire E. coli genome. Cell 9, 91-99.
Daly, J. M., 1981: Mechanisms of action. In: Durbin, R. D. (ed.), Toxins in Plant
Disease, pp. 331-394. New York: Academic Press.
Daniels, M. J., Barber, C. E., Turner, P. C., Cleary, W. G., Sawczyc, M. K., 1984:
Isolation of mutants of Xanthomonas campestris pv. campestris showing altered
pathogenicity. J. Gen. Microbiol. 130,2447-2455.
Day, P. R., 1974: Genetics of Host-Parasite Interaction, pp. 17-19. San Francisco:
W. H. Freeman and Company.
de Bruijn, F. J., Lupski, J. R., 1984: The use of transposon Tn5 mutagenesis in the
rapid generation of correlated physical and genetic maps of DNA segments
cloned into multicopy plasmids - a review. Gene 27, 131-149.
Dixon, L. K., Leaver, C. J., Brettell, R. I. S., Gengenbach, B. G., 1982: Mitochon-
drial sensitivity to Drechslera maydis T-toxin and the synthesis of a variant
mitochondrial polypeptide in plants derived from maize tissue cultures with
Texas male-sterile cytoplasm. Theor. Appl. Genet. 63, 75-80.
Federoff, N., 1983: Controlling elements in maize. In: Shapiro, J. (ed.), Mobile
Genetic Elements, pp. 1-63. New York: Academic Press.
Federoff, N., Furtek, D., Nelson, O. E., 1984a: Cloning of the bronze locus in maize
by a simple and generalised procedure using the transposable controlling
element Ac. Proc. Nat. Acad. Sci., U.S.A. 81, 3825-3829.
Federoff, N., Wessler, S., Shure, M., 1984 b: Isolation of the maize controlling ele-
ments Ac and Ds. Cell 35, 235-242.
Fincham, J. R. S., Sastry, G. R. K., 1974: Controlling elements in maize. Annu. Rev.
Genet. 8, 15-50.
Flor, H. H., 1947: Inheritance of reaction to rust in flax. J. Agric. Res. 74, 241-262.
Flor, H. H., 1956: The complementary genic systems in flax and flax rust. Adv.
Genet. 8, 29-54.
Forde, B. G., Leaver, C. J., 1980: Nuclear and cytoplasmic genes controlling syn-
thesis of variant mitochondrial polypeptides in male-sterile maize. Proc. Nat.
Acad. Sci., U.S.A. 77, 418-422.
Freeling, M., 1984: Plant transposable elements and insertion sequences. Annu.
Rev. Plant Physiol. 35, 277-298.
Gengenbach, B. G., Connelly, J. A., Pring, D. R., Conde, M. F., 1981: Mitochon-
drial DNA variation in maize plants regenerated during tissue culture selection.
Theor. Appl. Genet. 59,161-167.
Gheysen, G., Dhaese, P., Van Montagu, M., Schell, J., 1985: DNA flux across
genetic barriers: the crown gall phenomenon. In: Hohn, B., Dennis, E. S. (eds.),
Plant Gene Research: Genetic Flux in Plants, pp. 11-47. Wien - New York:
Springer.
Gilchrist, D. G., Grogan, R. G., 1976: Production and nature of a host-specific
toxin from Alternaria alternata f. sp. lycopersici. Phytopathology 66, 165 -171.
Hann, A. c., 1978: Studies on the epidemiology of Septoria species on wheat. Ph.
D. Thesis, Aberystwyth: University College of Wales.
Hartman, C. L., McCoy, T. J., Knous, T. R., 1984: Selection of alfalfa (Medicago
sativa) cell lines and regeneration of plants resistant to the toxins(s) produced
by Fusarium oxysporum f. sp. medicaginis. Plant Sci. Lett. 34, 183 -194.
Herrera-Estrella, L., Van den Broeck, G., Maenhaut, R., Van Montagu, M., Schell,
J., 1984: Light-inducible and chloroplast-associated expression of a chimaeric

gene introduced into Nicotiana tabacum using a Ti-plasmid vector. Nature
(London) 310, 115-l20.
Hilty, 1. W., Josephson, L. M., 1971: Reaction of corn inbreds with different cyto-
plasms to Helminthosporium maydis. Plant Dis. Rep. 55, 195-198.
Hooker, A. L., 1967: The genetics and expression of resistance in plants to rusts of
the genus Puccinia. Annu. Rev. Phytopathol. 5, 163-182.
Hooker, A. L., Smith, D. R., Lim, S. M., Beckett, J. B., 1970: Reaction of corn seed-
lings with male-sterile cytoplasm to Helminthosporium maydis. Plant Dis. Rep.
54,708-712.
Ingram, D. S., 1978: Cell death and resistance to biotrophs. Ann. Appl. BioI. 89,
291-295.
Inze, D., Follin, A., Van Lijsebettens, M., Simoens, C., Genetello, c., Van Montagu,
M., Schell, J., 1984: Genetic analysis of the individual T-DNA genes of Agro-
bacterium tumefaciens; further evidence that two genes are involved in indole-
3-acetic acid synthesis. Mol. Gen. Genet. 194, 265-274.
Kemble, R. J., Flavell, R. B., Brettell, R. I. S., 1982: Mitochondrial DNA analyses
of fertile and sterile maize plants derived from tissue culture with the Texas
male sterile cytoplasm. Theor. Appl. Genet. 62, 213-217.
Kent, S. S., Strobel, G. A., 1976: Phytotoxin from Septoria nodorum. Trans. Br.
Mycol. Soc. 67, 354-358.
Kono, Y., Knoche, H. W., Daly, J. M., 1981: Structure: fungal host-specific. In:
Durbin, R. D. (ed.), Toxins in Plant Disease, pp. 221-257. New York:
Academic Press.
Kuc, 1., 1972: Phytoalexins. Annu. Rev. Phytopath. 10,207-232.
Laughnan, J. R., Gabay-Laughnan, S., 1983: Cytoplasmic male sterility in maize.
Annu. Rev. Genet. 17,27-48.
Lawrence, G. J., Mayo, G. M. E., Shepherd, K. W., 1981: Interactions between
genes controlling pathogenicity in the flax rust fungus. Phytopathology 71,
12-19.
Leach, J., Tegtmeier, K. J., Daly, J. M., Yoder, O. c., 1982: Dominance at the ToxI
locus controlling T-toxin production by Cochliobolus heterostrophus. Physiol.
Plant Pathol. 21, 327-333.
Leaver, C. J., Gray, M. W., 1982: Mitochondrial genome organization and
expression in higher plants. Annu. Rev. Plant Physiol. 33, 373-402.
Leonard, K. J., 1983: Association of mating type and virulence in Helminthosporium
maydis. and observations on the origin of the race T population in the United
States. Phytopathology 63, 112-115.
Macko, V., 1983: Structural aspects of toxins. In: Daly, J. M., Deverall, B. J. (eds.),
Toxins and Plant Pathogenesis, pp. 41-80. Sydney: Academic Press Australia.
Manners, J. M., Davidson, A. D., Scott, K. J., 1985: Patterns of post-infectional
protein synthesis in barley carrying different genes for resistance to the
powdery mildew fungus. Plant Mol. BioI. 4, 275-283.
Miller, R. J., Koeppe, D. E., 1971: Southern corn leaf blight: Susceptible and
resistant mitochondria. Science 173,67-69.
Muller, K. 0., 1956: Einige einfache Versuche zum Nachweis von Phytoalexinen.
Phytopathol. Z. 27, 237 - 254.
Nelson, O. E., Ullstrup, A. J., 1964: Resistance to leaf spot in maize. J. Hered. 55,
195-199.
Niepold, F., Anderson, D., Mills, D., 1985: Cloning determinants of pathogenesis
from Pseudomonas syringae pathovar syringae. Proc. Nat. Acad. Sci., U.S.A. 82,
406-410.
246 R. I. S. Brettell and A J. Pryor
Nishimura, S., Kohmoto, K., 1983: Roles of toxins in pathogenesis. In: Daly, J. M.,
Deverall, B. J. (eds.), Toxins and Plant Pathogenesis, pp. 137-157. Sydney:
Academic Press Australia.
O'Reilly, C., Shepherd, N. S., Pereira, A., Schwarz-Sommer, Z., Bertram, I.,
Robertson, D. S., Peterson, P. A., Saedler, H., 1985: Molecular cloning of the al
locus of Zea mays using the transposable elements En and Mul. EMBO J. 4,
877-882.
Pereira, A., Schwarz-Sommer, Zs., Gierl, A., Bertram, I., Peterson, P. A., Saedler,
H., 1985: Genetic and molecular analysis of the Enhancer (En) transposable
element system of Zea mays. EMBO J. 4, 17-23.
Rufty, R. C., Herbert, T. T., Murphy, C. F., 1981: Variation in virulence in isolates
of Septoria nodorum. Phytopathology 71, 593-596.
Scheffer, R. P., 1983: Toxins as chemical determinants of plant disease. In: Daly, J.
M., Deverall, B. J. (eds.), Toxins and Plant Pathogenesis, pp. 1-40. Sydney:
Academic Press Australia.
Scheffer, R. P., Nelson, R. R., Ullstrup, A J., 1967: Inheritance of toxin production
and pathogenicity in Cochliobolus carbonum and Cochliobolus victoriae. Phyto-
pathology 57, 1288-1291.
Schroder, G., Waffenschmidt, S., Weiler, E. W., Schroder, J., 1984: The T-region of
Ti plasmids codes for an enzyme synthesising indole-3-acetic acid. Eur. J.
Biochem. 138,387-391.
Scott, P. R., Benedikz, P. W., Cox, C. J., 1982: A genetic study of the relationship
between height, time of ear emergence and resistance to Septoria nodorum in
wheat. Plant Patho!. 31, 45-60.
Shipton, W. A, Boyd, W. R. J., Rosielle, A A, Shearer, B. I., 1971: The common
Septoria diseases of wheat. Bot. Rev. 37, 231-262.
Staskawicz, B. J., Dahlbeck, D., Keen, N. T., 1984: Cloned avirulence gene of Pseu-
domonas syringae pv. glycinea determines race-specific incompatibility on
Glycine max (L.) Merr. Proc. Nat. Acad. Sci., U.S.A. 81, 6024-6028.
Sutton, W. D., Gerlach, W. L., Schwartz, D., Peacock, W. J., 1984: Molecular
analysis of Ds controlling element mutations at the Adhl locus of maize.
Science 223, 1265-1268.
Tavella, C. M., 1978: Date of heading and plant height of wheat varieties as related
to Septoria leaf blotch damage. Euphytica 27,577-580.
Torp, J., Andersen, B., 1982: Two-dimensional electrophoresis of proteins from cul-
tures of Erysiphe graminis f. sp. hordei. Physio!. Plant Patho!. 21, 151-160.
Welsh, J. N., Peturson, B., Machacek, J. E., 1954: Associated inheritance of
reaction to races of crown rust, Puccinia coronata avenae Erikss., and to Victoria
blight, Helminthosporium victoriae M. and M., in oats. Can. J. Bot. 32, 55-68.
Wynn, W. K., 1976: Appressorium formation over stomates by the bean rust
fungus: response to a surface contact stimulus. Phytopathology 66, 136-146.
Yoder, o. C., 1980: Toxins in pathogenesis. Annu. Rev. Phytopathol. 18, 103-129.
Note added in proof:
Further to the discussion on methods for identifying genes involved in patho-
genicity, Daniels et al. (1984) have isolated non-pathogenic mutants of Xan-
thomonas campestris pv. campestris following treatment of the bacteria with
chemical mutagen. As with the non-pathogenic strains of Pseudomonas syringae
obtained by transposon mutagenesis (Anderson and Mills, 1985; Niepold et aI.,
1985), it has yet to be established whether the mutations reside in the genes
involved in pathogen specificity.
Chapter 9
Gametophytic Gene Expression
David L. Mulcahy
Department of Botany, University of Massachusetts, Amherst, MA 01003, U.S.A.
Contents
I. Introduction
II. Overlap Between Sporophytic and Gametophytic Genotypes
III. Gametophytic Gene Expression and the Angiosperms
IV. The Influence of Haploid Genotype on Pollen Size
V. Time of Gene Expression in Pollen
VI. Methods of Haploid Selection
VII. Gametophytic Gene Expression and the Style
VIII. Gene Expression in the Megagametophyte
IX. Conclusions
X. References
I. Introduction
Gametophytic gene expression is a topic of both basic and applied interest.

It has been subjected to many reviews, starting with Buchholz (1922) and
continuing with unabated interest to the present (Zamir, 1983; Mulcahy,
1984). Consequently, this review will focus on general characteristics and
principles.
When a diploid organism, heterozygous at N loci, undergoes meiosis, it
has the potential of producing 2N different haploid genotypes. In the case
of pollen, this great variety is often associated with an equally impressive
overabundance of individuals. Insect-pollinated angiosperms, for example,
produce on the average about six thousand pollen grains for each ovule
and, in wind-pollinated species, the ratio averages one million to 1
(Cruden, 1977). The picture that these facts convey is one of a vast array of
pollen genotypes competing for a relatively small number of ovules. Cer-
tainly such competition holds the potential for extremely intense selection.
However, this potential will be realized only if genes are expressed in the
gametophyte. Futhermore, such pollen competition will effect the sporo-
phyte only if the genes which are expressed in the gametophyte are, at least
248 D. L. Mulcahy
in part, the same genes which are expressed in the sporophyte. Thus we
must ask if genes are expressed in the gametophyte, and, if so, are they
among the genes which are expressed in the sporophyte? Although sper-
matozoa in animals correspond to the sperm produced by the microgamet-
ophyte and comparisons of spermatozoa with pollen grains are misleading,
post-meiotic gene expression is apparently rare in animals (Braden, 1972).
Haldane (1932) speculated that it would be dis favoured by natural
selection in plants. He suggested that the potential intensity of pollen
selection could overwhelm selective values of the sporophyte and thus
allow the spread of gametophytically favoured, but sporophytically disfa-
voured, alleles. Selection should thus either reduce gene expression in the
pollen or limit it to genes which are not expressed in the sporophytic
portion of the life cycle. These concepts were apparently affirmed by the
fact that the gene waxy, expressed in the pollen of Zea mays, is not
expressed in the sporophyte (Nelson and Tsai, 1964). On the other hand,
Schwartz (1971) demonstrated conclusively that the Adh (alcohol dehy-
drogenase) loci of Zea mays are expressed in both phases of the life cycle.
Apparently examples can be found which do support Haldane's sugges-
tions and others which do not. A more general test of his ideas was needed,
not limited to single loci. The first such test was provided by Ter-Avanesian
(1949). He varied the intensity of pollen competition in Gossypium, Vigna
and Triticum by applying limited or excessive quantities of pollen to
stigmas. In the first case, fertilizations will be accomplished by gametes of
both fast and slow growing pollen tubes since ovules are not limiting. With
excessive pollinations, however, only the fastest growing pollen tubes will
reach unfertilized ovules. If pollen tube growth rate is determined by genes
which are expressed in the pollen, then it should be possible to select
among the segregating pollen genotypes of a single plant. Furthermore, if
the genes which are expressed, and thus subject to selection in the pollen,
are different from the genes which are expressed in the sporophyte, then
selecting among pollen tubes should have no effect upon the resulting
sporophytes. In each of the species he studied, Ter-Avanesian found that
the range of morphological variation exhibited in the resulting sporophytic
generation was greatest when limited amounts of pollen were used. Appar-
ently, genes are expressed in the pollen and many of these same genes are
expressed also in the sporophytic portion of the life cycle. A summary of
these and later studies was published by Ter-Avanesian (1978) and a
similar study was described by Matthews (see Lewis, 1954). Despite their
significance, these studies were largely overlooked.
Starting in 1975, further evidence was presented that the quality of the
next sporophytic generation is indeed influenced by the outcome of pollen
competition. For example, in Petunia hybrida, increasing the quantity of
pollen used in pollinations resulted in increased growth rates of Fl plants,
as indicated by the number of leaves produced and the total above-ground
fresh weight (Mulcahy et al., 1975). This study, and also the studies by Ter-
Avanesian, was complicated by the fact that varying the quantity of pollen
used in pollinations often modifies the number of seeds produced, which
Gametophytic Gene Expression 249
in turn introduces variation in the weights of seeds. However, at least in the

Petunia study, this added variable was apparently outgrown within 60 days
of planting. Furthermore, the influence of pollen selection persisted in the
plants resulting from selfing the Fl plants (Mulcahy et al., 1978). Never-
theless, another study was run in order to avoid altogether the possibility of
persistent maternal influences. The species used in this latter study was
Dianthus chinensis, chosen because its elongated stigmatic surfaces allow
pollination to take place at different distances from the ovary. This is
useful because the greater the distance travelled by a group of competing
pollen tubes, the greater will be the separation between the faster and the
slower individuals. In this way, the intensity of pollen competition can be
varied without modifying either the number of pollen grains applied or,
more importantly, the sizes of the resulting seeds. Again, in this study it
was found that pollen selection would modify the quality of the resulting
sporophytic generation. Specifically, sporophytes which resulted from the
gametes of the most rapidly growing pollen tubes showed significantly
greater speeds of germination and growth than did other sporophytes
(Mulcahy and Mulcahy, 1975). These studies cast serious doubts on
Haldane's suggestions about the nature (or lack) of gene expession in the
pollen. They also prompted a renewed interest in the extent to which spe-
cific genes were or were not expressed in the pollen.
II. Overlap Between Sporopbytic and Gametopbytic Genotypes
A series of studies has tested for the presumed overlap between haploid
and diploid expressed genotypes. The first of these was by Tanksley et aI.,
(1981). Pollen is known to contain many gene products but these could as
well be products of the diploid pollen source instead of the individual
pollen genotypes. To distinguish between these two alternatives, Tanksley
and associates utilized the fact that if an individual is heterozygous for a
gene encoding a dimeric enzyme, that individual will exhibit a gene
product unique to heterozygotes, the heterodimer. If the gene contents of
pollen are the products of sporophytic transcription and translation, pollen
of heterozygotes should contain heterodimers. Alternatively, if the pollen
contents are the results of postmeiotic gene expression, the pollen should
exhibit only homodimers. They examined thirty loci which code for
dimeric enzymes in Lycopersicon esculentum and closely related species.
They concluded that 60 % (18/30) of the genes which were expressed in the
sporophyte were expressed, without heterodimers, also in the pollen. This
was very strong evidence that a substantial fraction of the structural genes
which are expressed in the sporophyte of tomato are expressed also post-
meiotically. This conclusion would explain why selection in the pollen
could modify the subsequent sporophytic generation.
In view of the potential significance of this conclusion, Sari-Gorla et
al. (1983) tested the possibility that perhaps the absence of the hetero-
dimers in pollen could be due, not to postmeiotic gene expression but
250 D. L. Mulcahy
rather to something in the gametophytic cellular environment which pre-

vented the formation of heterodimers. Thus they constructed B-A translo-
cations in Zea mays in order to produce some pollen grains which would
carry two different alleles of selected loci. Functionally, this provided some
pollen which was partially diploid and heterozygous for a few loci, and
thus allowing an opportunity for the formation of heterodimers within the
pollen. Using this method, they concluded that some sporophytically
expressed loci, for example, glutamate-oxaloacetate-transaminase, got-I,
did form heterodimers in the pollen. Thus the absence of heterodimers in
Tanksley's study was, as they had concluded, due not to the inability to
form heterodimers but to postmeiotic gene expression.
Sari-Gorla and colleagues at the University of Milan have used the B-A
translocations also to test for overlap between sporophytically and gameto-
phytically expressed loci (Frova et al., in prep.). They also concluded that a
clear majority of the sporophytically expressed structural genes of Zea
mays are expressed also in the pollen (see Table 1).
Table I. Three Estimates of Overlap Between Pollen and Sporophytically

Expressed Genes.
Expressed in Expressed in both Expressed in Species Ref.

Pollen Only Pollen and Sporo- Sporophyte
phyte Only
3% 58 % (59 %) 39 % Lycopersicon Tanksley et aI.,

esculentum 1981
< 15 % 54 % 60 %) >31 % Tradescantia Willing et al.,
1984
6% 73 % (78 %) 21 % Zea mays Frova, (pers.
comm.)
Figures in parentheses indicate which percentage of the sporophytically expressed

genes are expressed in the pollen.
The most recent study of overlap between sporophytic and gametophytic

genomes involved a molecular survey and comparison of the mRNAs pro-
duced in the pollen and stems of Tradescantia paludosa (Willing and Mas-
carenhas, 1984). It thus differed from previous studies in the number of
genes which could be monitored. mRNA extracted from pollen and stems
was used to prepare cDNAs for those messages. Hybridization of homol-
ogous cDNA and RNA showed that pollen contained approximately
20,000 diverse mRNAs while shoots contained 30,000. This demonstrates
quite clearly that there is substantial gene activity in the pollen. Heterol-
ogous hybridizations, i. e., pollen cDNA to shoot poly(A)RNA and shoot
cDNA to pollen poly(A)RNA suggested that about 64 % of the pollen
sequences are similar to those in shoots and about 60 % of the sequences in
shoots are similar to those in pollen. These analyses were based on total
cellular poly(A)RNA, which allows the possibility that both tissues may
yield a great proportion of nuclear RNAs which mayor may not be trans-
lated. However, the authors point out that pollen is dehydrated at anthesis
and no RNA or protein synthesis is occurring. The use of total cell
poly(A)RNA seems justified also in the case of vegetative tissue, since
Galau et al. (1981) found in cotton that hybridization kinetics of DNA
complementary to total cell poly(A)RNA were virtually identical to the kin-
etics when polysomal poly(A)RNA was used.
Considering the above studies, summarized in Table 1, it seems
apparent that a large proportion of the structural genes expressed in the
sporophyte are expressed, and thus subject to selection, also in the pollen.
The sporophytic and gametophytic genotypes thus appear to overlap exten-
sively. This overlap may be significant since selection among haploid geno-
types is, in several situations, far more effective than selection among
diploid genotypes (Pfahler, 1983). For example, a rare recessive is immedi-
ately exposed to selection in a haploid genotype. Furthermore, multilocus
adaptations are more easily assembled in haploidy than in diploidy (Zamir,
1983). These considerations, and the fact that pollen is often produced in
very large numbers, mean that pollen selection is strictly comparable to the
mass selection which contributes so greatly to the adaptability of microbial
organisms. With overlap, these aspects of haploid selection could impinge
on substantial portions of the sporophytic genotype.
III. Gametophytic Gene Expression and the Angiosperms
Overlap between gametophytic and sporophytic genotypes brings the pot-

ential of haploid selection to the sporophytic portion of the life cycle.
However, this potential will be realized only to the extent that individual
gametophytes are able to complete with each other. Two unique features of
the angiosperms, evolved in response to other selective pressures (Crepet,
1984, for a review), serve to greatly accentuate pollen tube competition
(Mulcahy, 1979), namely insect pollination and the closed carpel. Insect
pollination increases the number of pollen grains reaching stigmatic sur-
faces, which should intensify pollen tube competition. Insect pollination
also means that pollen grains will tend to arrive in clumps, rather than as
individuals carried by the wind. This simultaneity of pollination reduces
the influence of random starting times and thus allows a more consistent
separation of faster from slower pollen tubes. The closed carpel increases
the distance through which pollen tubes must grow and, as discussed
earlier, this will improve the separation of faster and slower pollen tubes.
As a result, the consequences of gene expression in the gametophyte and of
genetic overlap may be greatly enhanced in the angiosperms.
252 D. L. Mulcahy
IV. The Influence of Haploid Genotype on Pollen Size
Johnson et al. (1976) found that pollen diameter in Zea mays was influ-
enced by both the sporophytic and the gametophytic genotypes. The sporo-
phytic influence was shown by a steady decline (Y = 98 + ( - 0.107) X) in
the average pollen grain diameter from the FI (x = 100.85 ll) to the F7
(x = 94.03 ll). Apparently, the inbred plants produce smaller pollen grains
than do the vigorous hybrids. Superimposed on this sporophytically deter-
mined decline in average diameter is a gametophytic influence. This re-
flects the fact that inbreeding reduces heterozygosity and the pollen geno-
types thus become more homogeneous as inbreeding progresses. As a
result, the coefficient of variation in pollen diameter also declines with
inbreeding (r = -0.5530, P < 0.01). Variance in pollen diameters reflects
not only the influence of the genes carried within the individual pollen
grains but also competitive gametophytic interactions. A pollen lethal
allele on the fourth chromosome of Zea mays greatly reduces the diameter
of pollen grains carrying it (Singleton and Mangelsdorf, 1940). Fur-
thermore, the maximum diameter of the wild-type segregants is clearly
greater than that of plants not segregating for the pollen lethal, indicating
that in the segregating plants the wild-type gametophytes are able to
pre-empt resources relinquished by those carrying the pollen lethal factor.
It is not known if these greater resources allow the functional grains any
competitive advantages either through greater longevity or increased pollen
tube growth rates.
V. Time of Gene Expression in Pollen
Stinson and Mascarenhas (1985) found that no ADH activity was detect-
able when tetrads of Zea mays were first formed. Activity was first detected
soon after the tetrads began to break apart but before separation was com-
plete. In Medicago sativa, however, glutamate dehydrogenase activity first
appeared in the mitochondria of pollen of the binucleate stage, that is,
close to pollen maturity (Vienne et al., 1985). Apparently, different
enzymes become active at different stages of pollen development. A syste-
matic investigation of this subject could suggest which enzymes are
essential for pollen development and which function during pollen tube
growth. This information would ultimately provide useful information in
programmes designed to select particular qualities by means of pollen
selection.
VI. Methods of Haploid Selection
Considering the microgametophytic portion of the life cycle, two subsets of

selection phenomena should be considered: micros pore selection and
pollen selection. The first of these includes all selection occurring between
the completion of meiosis and that of pollen maturity. Pollen selection, in

contrast, extends from pollen maturity until zygote formation. In all cases,
only phenomena which are controlled by genes transcribed and translated
after meiosis should be considered.
Zamir and Vallejos (1983) contrasted the relative effectiveness of the
two subsets of microgametophytic selection in Lycopersicon spp. In one
case, ramets of an F1 between Lycopersicon esculentum and a cold tolerant
accession of L. hirsutum were maintained at controlled environments of
either 24/19 C and 12/6 and then backcrossed to Lycopersicon escu-
lentum grown in a greenhouse (approximately 24/18 C). In another case,
pollen from a greenhouse maintained F1 clone was backcrossed to Lyco-
persicon esculentum in the controlled environments. In the first case, any
differential effects of the two controlled environments would be applied
only during microspore development. In the second case, differential was
applied during pollen germination, pollen tube growth, and fertilization.
In each treatment, 8 isozyme loci were analyzed in the backcross
progeny to determine if either Lycopersicon esculentum or L. hirsutum loci
deviated from the expected 1: 1 ratio. When controlled temperatures were
applied to the microspores, one locus, Aps-2, deviated significantly with the
higher temperature, Got-3 and Est-4 deviated with the lower temperature,
and two loci, Pgm-2 and Pgi-l, deviated with both temperature regimes. In
each of these five cases, the significant deviations were in favour of the L.
esculentum isozyme. When the different controlled temperatures were
applied to the pollen, significant deviations in favour of the L. esculentum
isozymes (Aps-2, Pgi-l and Est-4) occurred at the higher temperature. At
the lower temperature, significant deviations occurred in favour of L.
hirsutum isozymes (Adh-2, Pgi-l and Est-4). From these data, it is clear that
selection for isozymes associated with cold tolerance was much more
effective with pollen than with microspores.
In contrast to the clear superiority of pollen selection observed by Za-
mir and Vallejos (1983), Searcy and Mulcahy (1985) found that microspore
selection was the more effective means of selecting for tolerance to heavy
metals. Working with metal-tolerant and metal-nontolerant ecotypes of
Silene dioica and Mimulus guttatus, they produced hybrids between tolerant
and nontolerant parents segregating for metal tolerance. When the F1
plants were grown in the presence of copper, in the case of Mimulus, and
zinc, for Silene, the proportion of potentially functional pollen was reduced
to 54.0 % and 61.4 % of the control values, respectively. This by itself sug-
gests that the hybrids are each producing two subpopulations of pollen,
one tolerant and one nontolerant to the heavy metal. When backcrossed to
nontolerant pistillate parents, pollen of segregating plants of Mimulus and
Silene, normally produced 32.9 % and 54.2 % tolerant progeny, respectively.
However, if the segregating pollen sources are treated with heavy metal
during pollen development, the proportion of tolerant progeny resulting
from backcrosses increases to 59.1 % in the case of Mimulus and to 79.2 %
in the case of Silene. This amounts to a 79.6 % increase in the proportion of
tolerant progeny in the case of Mimulus and a 46.1 % increase with Silene.
254 D. L. Mulcahy
The efficacy of pollen selection, as opposed to microspore selection,

was tested by growing pistillate plants which were tolerant to heavy metals
either with or without heavy metals. Growing Mimulus in nutrient solutions
containing 0, 1 or 2 ppm copper resulted in 6, 23, and 48 ppm copper in the
flowers, respectively. For Silene, growth on 0 and 5 ppm zinc resulted in
34-40 and 150-200 ppm zinc within the styles, respectively. Pollen for
backcrosses to these plants was obtained from tolerant and nontolerant
plants grown without heavy metals. The test of pollen selection was to
measure the growth rates of pollen tubes in the treated and untreated
styles. In Silene, neither the tolerance of the pollen source nor the concen-
tration of heavy metal within the style influenced pollen tube growth rates.
With Mimulus, pollen from the nontolerant pollen source grew signifi-
cantly faster than did that from the tolerant source. However, here too, the
concentration of heavy metal in the style failed to have a significant effect
upon pollen tube growth rate.
For whatever reason, it appears that microspore selection is far more
effective in the case of heavy metals and that pollen selection is more
effective for isozymes which are associated with cold tolerance.
An alternative method of selecting among pollen genotypes is to treat
the mature pollen with in vitro techniques. Schwartz and Osterman (1976)
took advantage of the fact that alcohol dehydrogenase (ADH) converts
allyl alcohol to acrylaldehyde, a highly toxic compound, to select for ADH
mutants. Pollen from Zea mays which lacks alcohol dehydrogenase sur-
vives exposure to allyl alcohol whereas wild-type pollen is killed.
An additional if undirected form of selecting pollen grains is to subject
them to storage. Pfahler (1967) found that pollen from separate sporo-
phytes of Zea mays showed significant differences in the ability to tolerate
storage. This could reflect sporophytic as well as gametophytic genotypes.
Later investigations, however, indicated that specific alleles from a segre-
gating plant showed significant effects on storage tolerance (Pfahler et al.,
1981). Furthermore, in Petunia hybrida, the pollen grains which are best
able to survive storage give rise to sporophytes which grow significantly
more rapidly than do sporophytes resulting from unselected pollen geno-
types (Mulcahy et al., 1982). This last result may seem rather unexpected
until considered in light of the Singleton and Mangelsdorf (1940) study,
which showed that more vigorous gametophytes are able to pre-empt
resources from others within the same anther locule. Perhaps such pre-
emption is associated with a greater ability to tolerate storage and also con-
tributes to increased growth rates in the resultant sporophytic generation.
VII. Gametophytic Gene Expression and the Style
In all of the above studies, the style is either ignored or considered to be a

passive arena within which microgametophytic gene expression is mani-
fested. This convenient assumption neglects extensive ~vidence that the
style has a very significant influence on pollen and pollen gene expression.
Self-incompatibility is the most widely recognized expression of pollen-

style interactions but there are others. In Zea mays, the stylar genotype has
a significant influence on pollen tube growth rate. Furthermore, there are
significant interactions between pollen and stylar genotypes, meaning that
not all pollen types are affected in the same way (Pfahler, 1967; Sari-Gorla
and Bellintani, 1976). Within the style there are also interactions between
different pollen tubes (Sari-Gorla and Rovida, 1980).
VIII. Gene Expression in the Megagametophyte
Owing to the facts that ovules are both less numerous and less accessible
than pollen grains, the megagametophyte has received relatively little
attention. However, Schwemmle (1968) and colleagues have demonstrated
there too that genes are expressed. Using the well known Renner com-
plexes in Oenothera spp., they showed for several cases that the excess of
heterozygotes usually obtained was not due to abortion of pollen grains or
embryo sacs which carry one complex or the other. Rather the excess
results from selective fertilization; the tendencies for pollen tubes and
ovules to unite is determined by the Renner complex carried in each. It is
particularly significant that selective attraction between different ovule and
pollen types could be demonstrated also in vitro. Grant (1973) quotes
studies by Lamprecht indicating that selective fertilization occurs also in
Pisum sativum. It would seem that with the ready availability of isozyme
markers, selective fertilization and megagametophytic gene expression are
now accessible for study.
IX. Conclusions
Studies of several angiosperm species have indicated that many structural

genes are expressed in the pollen. By one estimate, the number of such
genes is approximately two thirds the number of structural genes expressed
in the sporophyte. Genes expressed in the microgametophyte influence
competitive interactions between developing microspores, pollen diameter,
pollen germination, pollen tube growth rate, interactions between pollen
tubes, and nonrandom fertilization. A large fraction (64-95 %) of micro-
gametophytically expressed genes are expressed also in the sporophyte.
Thus, up to 60 % of the sporophytic structural genes are exposed to
selection in the pollen. This extensive overlap between gametophytic and
sporophytic genomes explains why pollen selection influences sporophytic
qualities such as seed germination, plant growth rate, competitive ability,
yield, and tolerance to heavy metals, as well as other forms of stress. The
efficacy of pollen selection stems largely from the fact that pollen is
haploid and available in large populations. It thus provides a functional
equivalent of the mass selection which contributes so greatly to the adapt-
ability of microorganisms. The angiosperms possess characteristics which
256 D. L. Mulcahy
accentuate pollen selection and this may have contributed to the adapt-
ability and success of that group. Megagametophytic gene expression has
been less studied but data available suggest it too may be an arena of
extensive gene expression and selection.
Acknowledgements
Sincere thanks are due to my collaborators, Gabriella Bergamini Mulcahy

and Douglas MacMillan, for many helpful discussions in the preparation
of this paper. Our efforts are supported, in part, by USDA Biotechnology
Grant No. 8502969.
Dedication
This manuscript is dedicated to Prof. H. F. Linskens. The breadth and sig-
nificance of his many contributions to pollen biology stand as an
admirable model for us all. I look forward to the continuation of his pro-
ductivity for many years to come.
x. References
Braden, A. W. H., 1972: T-Locus in mice: segregation distortion and sterility in the
male. In: Beatty, R. A., Glueckshon-Waelsch, S. (eds.), Proc. Int. Symp. The
Genetics of the Spermatozoon, pp. 289-305. Edinburgh: University of Edin-
burgh.
Buchholz, J. T., 1922: Developmental selection in vascular plants. Bot. Gaz.
(Chicago) 73, 249-286.
Crepet, W. L., 1984: Advanced (constant) insect-pollination mechanisms: pattern of
evolution and implications vis-a-vis angiosperm diversity. Ann. Mo. Bot. Gerd.
71, 607 -630.
Cruden, R. W., 1977: Pollen-ovule ratios: a conservative indicator of breeding
systems in flowering plants. Evolution 31, 32-46.
Galau, G. A., Legocki, S. C., Greenway, S. C., Dure, L. S., 1981: Cotton messenger
RNA sequences exist in both polyadenalated and non-polyadenalated forms. J.
BioI. Chern. 256, 2552-2560.
Grant, V., 1973: Genetics of flowering plants, p. 239. New York: Columbia Uni-
versity Press.
Haldane, J. B. S., 1932: The Causes of Evolution. London: Longmans, Green and
Co.
Johnson, C. M., Mulcahy, D. L., Galinat, W. c., 1976: Male gametophyte in maize:
influence of the gametophytic genotype. Theor. Appi. Genet. 48, 299-303.
Lewis, D., 1954: Annual report of the department of genetics. Annu. Rep. John
Innes Hort. Inst. 45, 12-17.
McKenna, M., Mulcahy, D. L., 1983: Ecological aspects of gametophytic compe-
tition in Dianthus chinensis. In: Mulcahy, D. L., Ottaviano, E. (eds.), Pollen:
Biology and Implications for Plant Breeding, pp. 419-424. New York:
Elsevier.
Mulcahy, D. L., Robinson, R. W., Ihara, M., Kesseli, R., 1981: Gametophytic tran-
scription for acid phosphatases in pollen of Cucurbita species hybrids. J. Hered.
72,353-354.
Mulcahy, D. L., 1979: The rise of the angiosperms: a genecological factor. Science
206,20-23.
Mulcahy, D. L., 1984: Manipulation of gametophytic populations. In: Lange, W. et
al. (eds.), Efficiency in Plant Breeding, pp. 167-175. Wageningen: Pudoc.
Mulcahy, D. L., Mulcahy, G. B., 1975: The influence of gametophytic competition
on sporophytic quality in.Dianthus chinensis. Theor. Appl. Genet. 46, 277-280.
Mulcahy,D. L., Mulcahy, G. B., Ottaviano, E., 1975: Sporophytic expression of
gametophytic competition in Petunia hybrida. In: Mulcahy, D. L., Ottaviano, E.,
(eds.), Pollen: Biology and Implications for Plant Breeding, pp. 227-232. New
York: Elsevier.
Mulcahy, D. L., Mulcahy, G. B., Ottaviano, E., 1978: Further evidence that,gameto-
phytic selection modifies the genetic quality of the sporophyte. Soc. bot. Fr.,
Actualites botaniques 1978, 57 -60.
Mulcahy, G. B., Mulcahy, D. L., Pfahler, P. L., 1982: The effect of delayed polli-
nation in Petunia hybrida. Acta Bot. NeerI. 31, 97-103.
Nelson, O. E., Tsai, C. Y., 1964: Glucose transfer from adeninosine diphosphate-
glucose in preparations of waxy seeds. Science 145, 1194-1195.
Ottaviano, E., Sari-GorIa, M., Mulcahy, D. L., 1980: Pollen tube growth rates in
Zea mays: implications for genetic itnprovement of crops. Science 210,
437-438.
Pfahler, P. L., 1967: Fertilization ability of maize pollen grains II. Pollen genotype,
female sporophyte, and pollen storage interactions. Genetics 57, 513-521.
Pfahler, P. L., 1983: Comparative effectiveness of pollen genotype selection in
higher plants. In: Mulcahy, D. L., Ottaviano, E. (eds.), Pollen: Biology and
Implications for Plant Breeding, pp. 361-366. New York: Elsevier.
Pfahler, P. L., Linskens, H. F., Mulcahy, D. L., 1981: Effect of pollen ultraviolet
radiation on the abortion frquency and segregation patterns at various endo-
sperm mutant loci in maize (Zea mays L.). Environ. Exp. BioI. 21, 5-13.
Sari-GorIa, M., Bellintani, R., 1976: Variation in pollen fertilizing ability in relation
to the genotype of the stylar tissue. Maize Genet. News 50, 77 - 79.
Sari-GorIa, M., Rovida, E., 1980: Competitive ability of maize pollen: intergameto-
phytic effects. Theor. Appl. Genet. 57, 37-42.
Sari-GorIa, M., Frova, C. V., Ottaviano, E., Soave, c., 1983: Gene expression at the
gametophytic phase in maize. In: Mulcahy, D. L., Ottaviano, E. (eds), Pollen:
Elsevier.
Schwartz, D., 1971: Genetic control of alcohol dehydrogenase - a competition
model for regulation of gene action. Genetics 67,411-425.
Schwartz, D., Osterman, J., 1976: A pollen selection system for alcohol dehy-
drogenase-negative mutants in plants. Genetics 83, 63-65.
Schwemmle, J., 1968: Selective fertilization in Oenothera. Adv. Genet. 14,225-324.
Searcy, K., Mulcahy, D. L., 1985: Selection for heavy metal tolerance during pollen
development among pollen grains from a single individual. Amer. J. Bot. 72,
1700-1706.
Singleton, W. R., Mangelsdorf, P. C., 1940: Gametic letha1s on the fourth chromo-
some of maize. Genetics 25, 366-390.
Stinson, J., Mascarenhas, J. P., 1985: Onset of alcohol dehydrogenase synthesis
during microsporogenesis in maize. Plant Physiol. 77, 222-224.
258 D. L. Mulcahy
Tanksley, S., Zamir, D., Rick, C. M., 1981: Evidence for extensive overlap of sporo-
phytic and gametophytic gene expression in Lycopersicon esculentum. Science
213, 453-455.
Ter-Avanesian, D.,4-949: The influence of the number of pollen grains used in pol-
lination. Bull. Appl. Genet. Plant Breeding Leningrad. 28, 119-133.
Ter-Avanesian, D., 1978: The effect of varying the number of pollen grains used in
fertilization. Theor. Appl. Genet. 52, 77-79.
Vienne, D., Savina, F., Daussant, J., 1985: La glutamate deshydrogenase au cours
de la gametogenese miile chez Medicago sativa. Physiol. Plant. 63, 208-214.
Willing, R. P., Mascarenhas, J. P., 1984: Analysis of the complexity and diversity of
mRNAs from pollen and shoots of Tradescantia. Plant Physiol. 75, 865-868.
Zamir, D., Vallejos, E. C., 1983: Temperature effects on haploid selection of tomato
microspores and pollen grains. In: Mulcahy, D. L., Ottaviano, E. (eds.), Pollen:
Elsevier.
Zamir, D., 1983: Pollen gene expression and selection: applications in plant
breeding. In: Tanksley, S. D., Orton, T. J. (eds.), Isozymes in Plant Genetics and
Breeding. Part A., pp. 313-330. Amsterdam: Elsevier.
Chapter 10
Auxotroph Isolation In Vitro
Anne D. Blonstein
Friedrich Miescher-Institut, P. O. Box 2543, CH-4002 Basel, Switzerland
Contents
I. Introduction
A. Some Preliminary Comments
B. Why Auxotrophs?
C. The Characteristics, Advantages and Disadvantages
of In vitro Systems for Mutant Isolation
D. Species Suitable for Auxotroph Isolation
II. Mutagenesis
III. Methods of Selection
A. Conditions and Media
B. Positive Selection
C. Negative Selection
IV. Phenotypes
A. Amino-Acid, Vitamin and Purine Auxotrophs
B. Nitrate Reductase Mutants
C. Temperature-Sensitive Mutations
V. Fusion and Transformation
VI. Conclusions and Future Prospects
VII. References
I. Introduction
A. Some Preliminary Comments

Following the first report of a nutritional mutant in higher plants (Lang-
ridge & Brock, 1961), just ten years elapsed before auxotrophic mutants
were declared to have been produced from plant somatic cell culture
(Carlson, 1970), and the future looked bright for the application of in vitro
techniques in the induction and isolation of "novel" plant variation. Never-
theless, in 1980 Robert Shields, reviewing a book on plant cell and tissue
culture, commented that there is a "paucity of useful selectable mutations
in defined genes in plant cells" even though with the availability of haploid
material "recessive mutations should be obtainable at reasonable fre-
260 A. D. Blonstein
quency". He went on to say, "one can only assume that not enough people
have tried hard enough."
This review, concentrating on auxotrophic mutants (which can be
regarded as selectable in culture in that they can be selected against), will
examine the accuracy of this statement. It will show that much time and
effort has been put into in vitro mutant isolation and when problems and
setbacks have been encountered some have been solved, others not. There
has been both success and failure.
However, it will also highlight some of the approaches that seem to
have been neglected or under-exploited in this field. One of the penalties
of scientific specialization has been marked lack of co-operation and col-
laboration between tissue-culture technologists on the one hand and
biochemists, physiologists, and developmental biologists on the other. A
better awareness by the former of "need" and by the latter of available
techniques could result in some significant advances in knowledge and
understanding in key areas of plant biology.
B. Why Auxotrophs ?
There is an extensive literature on the role auxotrophic mutants have
played in the elucidation of biochemical processes in micro-organisms.
Enzyme lesions may lead to the accumulation of precursors which can be
identified directly or by the feeding of radiolabelled compounds. Altered
enzymes themselves can be extracted and analysed. Genetic complemen-
tation studies between phenotypically similar mutants will establish the
complexity of the system under study and can ultimately lead to an idea of
the physical organization and distribution of the related genes within the
genome. In addition to biosynthetic lesions, nutritional mutants may show
alterations in uptake, transport, degradation etc. Auxotrophs are also a
powerful tool for gene isolation via transformation.
Although nutritional mutants were found with comparative ease in bac-
teria and fungi, they proved somewhat more difficult to isolate in simple
lower plants, but success has been reported in Chlamydomonas (require-
ments for carbon sources, nicotinamide, para-aminobenzoic acid (PABA),
thiamine and purines - Gowans, 1960), Physcomitrella patens (thiamine,
PABA, niacin and yeast extract auxotrophs - Engel, 1968), and the liv-
erwort Sphaerocarpos donnelli (nicotinic acid, choline, glucose and arginine
requiring - Schieder, 1976).
Until the introduction of in vitro techniques, higher plants proved even
more recalcitrant, and auxotrophs were available for only a very restricted
range of compounds, and were frequently leaky: tomato (thiamine -
Langridge and Brock, 1961); Arabidopsis (thiamine: over 200 independent
mutations at at least 4 loci - Feenstra, 1964; Redei, 1965, 1975; Li and
Redei, 1969), Plantago insularis (thiamine - Murr and Spurr, 1973; Murr
and Stebbins, 1971), Zea mays (proline - Gavazzi et al., 1975; Racchi et
al., 1978).
The various reasons advocated for this apparent inability to isolate
Auxotroph Isolation In Vitro 261
plant auxotrophs by conventional seed mutagenesis are worth reiterating

because they both justify the use of in vitro technology and highlight com-
plications that even tissue culture cannot resolve.
Feenstra (1964) predicted that repairable recessive mutants would only
be found for compounds non-essential for male gamete development and
seed formation or, if essential, for ones which can be supplied by the
maternal tissue or receiving pistil and diffuse to the pollen or seed. Mutant
plants themselves must be able to take up the nutrient from the substrate
(or through their leaves, perhaps, if they are sprayed) and transport it to the
locations where it is needed. Nelson and Burr (1973) again emphasized the
failure of most auxotrophic mutants to survive embryonic development
when borne on heterozygous, phenotypically normal plants.
In addition, Li et al. (1967) speculated that auxotrophs may be rare in
green plants because a) their genetic material may be extensively dupli-
cated, b) alternative metabolic pathways co-exist and c) basic biochemical
reactions may be coupled to photosynthesis or other photochemical pro-
cesses.
C. The Characteristics, Advantages and Disadvantages of In vitro Systems

for Mutant Isolation
Although they have been noted many times before, it is important to review
and comment on the factors which have encouraged the use of cell-culture
systems for the induction of biochemical mutants.
First and foremost, a very large (i. e. several million) number of
genomes can be subjected to mutagenesis and screened for a desired trait.
In the case of recessive mutations (and most auxotrophs can be expected to
be recessive) as long as a chimera is not produced, selection can be applied
in culture and regenerated plants will be homozygous recessive. There will
be no need to wait until an F 2 generation to see segregation of the mutant.
Manipulations to "rescue" an auxotroph might be quite easy in the cell
culture environment, but feeding problems may still arise at and after plant
regeneration: the trait may not be sexually transmitted due to gametic
lethality. It may have to be accepted that some traits can only be studied in
culture (but see IV. B). In such cases it may be very difficult to distinguish
epigenetic from mutational events.
Freshly isolated leaf protoplasts are the preferred source of material for
mutagenesis and subsequent culture, selection and regeneration. Estab-
lished cell cultures have a variety of disadvantages: their regeneration
capacity is usually very low, possibly due to an accumulation of chromo-
somal alterations, mutations and biochemical (especially hormonal) "con-
ditioning"; cells aggregate; and they are difficult to synchronize. In con-
trast, leaf-derived protoplasts are usually highly synchronous and "frozen"
in G 1 at the time of isolation (Magnien et al., 1980; Magnien and Devreux,
1980; Galbraith et al., 1983).
The availability of haploid plants as a source of protoplasts was further
impetus in the search for auxotrophic mutants because only haploid
262 A. D. Blonstein
material allows the immediate expression of recessive mutations induced

early in culture. Most haploid cultures diploidize quite rapidly. Although
stable haploid cultures have been reported (e. g. Furner et al., 1978), it is
questionable whether they possess any advantages over haploid leaf-
derived protoplasts, and are subject to all the drawbacks of continuous
culture that have already been mentioned.
D. Species Suitable for Auxotroph Isolation

About eight species, most of them Solanaceae, are currently used, or
probably will be in the very near future, as in vitro systems for biochemical
mutant isolation. These are described briefly below.
i) Nicotiana tabacum
The first plant species for which in vitro culture from protoplast to plant
was reported (Nagata and Takebe, 1971). Its advantages as an experi-
mental system include well-established genetics and cytogenetics, and very
high regeneration capacity. However, as an allotetraploid possibly con-
taining the duplication of many genes it is less than ideal as starting
material when screening for recessive mutations, although it should be
added that these have been found.
ii) Nicotiana sylvestris

One of the two true diploid Nicotiana species that have been used as alter-
natives to tobacco. Although protocols have been reported for the culture
from protoplasts to plants (Nagy and Maliga, 1976; Durand, 1979; Bourgin
et al., 1979) there continue to be various reports of manipulation problems
with this species (e. g. Magnien and Devreux, 1980).
iii) Nicotiana plumbaginifolia

Currently one of the most popular "model" species and a true diploid, for
which sterile haploid shoot cultures are easily maintained. The passage
from protoplast to flowering plant takes about sixteen weeks (Bourgin et
al., 1979; N egrutiu, 1981), but in contrast to tobacco, N. plumbaginifolia
appears to lose its regeneration capacity quite rapidly in culture and regen-
erants are frequently infertile or self-sterile. It may be that true diploids,
spending a period of time in culture as haploids, may be more susceptible
and sensitive than polyploid species to mutagenesis and factors generating
somac1onal variation.
iv) Petunia hybrida

Haploid plants are available in this genetically well characterized species,
and plant regeneration from haploid and diploid protopiasts has been
reported (Durand et al., 1973; Binding, 1974).
v) Datura innoxia
In use in some laboratories (Schieder, 1975).
vi) Hyoscyamus muticus

A diploid species for which a protocol for culture of haploid protoplasts to
plants was reported (Lorz et al., 1979). Since then it has very successfully
been used for auxotroph isolation in vitro but fertile plant regeneration of
both mutants and wild type has been very difficult to obtain to date.
vii) Solanum tuberosum

Although di- and mono-haploids have been available in potato for some
years (Lamm, 1938; Dunwell and Sunderland, 1973; van Breukelen et aI.,
1975; Sopory et al., 1978) and techniques for protoplast culture and plant
regeneration are well established (Behnke, 1976; Binding et al., 1978;
Thomas, 1981), many cultivars are often not sexually fertile, a drawback to
their application as a model genetic system (Shepard, 1980).
viii) Lycopersicon esculentum

Tomato would be an ideal species for in vitro work having a very well-
defined genetic system. In addition, biochemical markers already exist at
the whole plant level which may express in culture. However, as yet,
haploid material is not generally available, and the techniques for the
regeneration of fertile plants from protoplasts of 14 diverse cultivars
reported recently (Shahin, 1985) remain to be shown to be universally
applicable. Nevertheless, tomato is set to become a very useful experi-
mental system in the near future.
II. Mutagenesis
The mutagenesis of plant cells - its theory and practical methodology -

has recently been reviewed in detail (Blonstein and King, 1985) and this
section will provide just a brief overview of the subject.
Mutagens should only be applied to plant cells if it seems absolutely
necessary. When a mutagen is applied, the generation of undesirable sec-
ondary mutations, and the (probably associated) frequently observed
reduction in plant regeneration capacity (Negrutiu et al., 1983) may often
outweigh the benefits of increased mutation frequency. However, if the
spontaneous frequency of a desired mutation is low, in the absence of an
effective selection scheme (see below) mutagenesis will probably be opted
for to reduce the effort involved in screening.
Cell irradiation (by UV, y- or X-rays) is preferable to the use of chemical
mutagens if only because less culture manipUlation (centrifugation, washing
etc.) is required. However, both irradiation and chemicals are believed to
have been successful in inducing mutants (Blonstein and King, 1985).
264 A. D. Blonstein
Low killing rates are desirable to preserve regeneration capacity, and

have been shown to be associated with an adequately high mutant fre-
quency. Indeed, Colijn et ai. (1979) only found drug-resistant calli at nitro-
soguanidine concentrations that had no observable effect on the survival of
mutagenized cultures.
One further factor of great importance is the timing of mutagen appli-
cation. There are two aspects to this problem: First, the stage(s), in the case
of synchronized cultures, in the cell cycle which are most mutagen sus-
ceptible, and second, the necessity to generate pure clones i. e. to ensure
that the DNA strands which segregate into daughter cells both carry the
mutation. Very little is known about the precise biochemical modes of
action of mutagens in plant cells and information must, perforce, be
extrapolated from the whole plant and microorganism literature.
Some mutagens (e. g. MNNG) seem to be most effective on replicating
DNA (Gebhardt et ai., 1981) and, in the case of mesophyll-derived proto-
plasts, maximal DNA replication usually occurs 1 to 2 days after isolation
and is paralleled by an increase in irradiation sensitivity (Galun and Raveh,
1975; Magnien et ai., 1980). However, there are few data for plant cells esta-
blishing that mutation frequency is definitely linked to DNA replication.
Additionally, mutations induced on one strand of replicating DNA and
fixed post-replicatively will result in chimeras which will probably express
as wild types when subjected to selection for auxotrophy. The ideal
mutagen causes lesions that are fixed in a pre-replicative process and will
probably, therefore, have to be active on non-replicating DNA. The infor-
mation simply is not available for plant cells as to which mutagens
promote such mechanisms. In work with yeast (N asim et ai., 1981), UV and
methyl methane sulphonate produced mainly pure clones, while EMS and
nitrous acid produced mainly mosaics.
III. Methods of Selection
A. Conditions and Media

Several important factors have to be taken into consideration when
designing a selection system for auxotroph isolation. The "true" wild-type
phenotype must first be established and tissue culture artifacts must be
recognized when choosing the base line from which to look for auxotrophs.
In a search for amino acid and vitamin auxotrophs, Gebhardt et al. (1981)
used one white and one green wild-type line, which differed somewhat in
their growth properties, as a means to reducing the potential number of
"false" positives selected in the early stages of their experiments.
While a whole plant may be prototrophic for a given compound, it may
not be required and/or synthesized by undifferentiated tissue growing in
culture. If a search is to be made for fairly specific mutations, it may be
wise to establish the presence of the compound in wild-type tissue before
screening for auxotrophs.
It has long been recognized that the composition of the "rescue"

medium can profoundly influence the phenotypes isolated. Gresshof
(1978) pointed out several problematic factors such as antagonism of
metabolites at high concentrations, or their supply in insufficient amounts.
In early work with Neurospora Lein et al. (1948) found that three of seven
auxotrophic mutants would not grow on complete medium. Histidineless
mutants were always inhibited if certain other amino acids were present in
the medium (Haas et al., 1952). Isolating auxotrophs in a liverwort,
Schieder (1976) also observed that the medium may influence what is iso-
lated.
The choice of concentration of a given supplement in the medium may
favour the isolation of certain kinds of mutants, but hinder the appearance
of others. Thus dose response curves for histidine- and tryptophan- vari-
ants of Hyoscyamus muticus show optima very close to the concentrations
used in the rescue medium for the original isolation experiments (Gebhardt
et al., 1983; Shimamoto and King, 1983).
Mutant isolation schemes frequently result in low density growth condi-
tions either as a deliberate intention, e. g. to prevent coalescence of
adjacent colonies before dilution, to limit cross feeding during enrichment
(see below), and to limit selection pressures against auxotrophs growing
more slowly than prototrophs in mixed culture (Roth and Lark, 1982), or as
the inevitable consequence of high kill after mutagenesis. Many cultures do
not survive at cell densities less than about 104 ml- I, and various methods
have been developed to either supply nutrients that may be leaking from
viable cells into the medium, or to counteract putative toxic products
released by dead or dying cells (although these could quite plausibly act as
a feeder themselves). Advocated methods include feeder layers (Raveh et
al., 1973; Weber and Lark, 1979; Shneyour et al., 1984; Horsch and King,
1984, 1985) and the use of enriched medium (Kao and Michayluk, 1975).
B. Positive Selection
Undoubtedly the easiest way to screen a population of cells for a mutant
phenotype is to apply positive selection, i. e. design a scheme which kills
wild-type cells and allows only mutants to survive. Theoretically, a vast
number of genomes can be screened this way. Unfortunately, this is not an
obvious technique for detecting auxotrophs - the auxotroph would need
to express an "advantage" over the wild type under a given set of condi-
tions, in addition to its growth requirement. So far in the field of in vitro
auxotroph isolation there is only one phenotype where this technique has
been fully exploited and yielded a large number of mutants at several loci:
nitrate reductase-deficient (NR -) mutants which cannot utilise nitrate as
their sole nitrogen source but exhibit resistance to chlorate because they
are unable to convert it to toxic chlorite as the wild type does (Aberg,
1947). Using this screen, NR - mutants and variants have been isolated in
vitro in tobacco (Muller and Grafe, 1978; Buchanan and Wray, 1982),
Datura innoxia (King and Khanna, 1980), Rosa damascena (Murphy and
266 A. D. Blonstein
Imbrie, 1981) and Nicotiana plumbaginifolia (Marton et al., 1982; Negrutiu

et al., 1983). Chlorate resistance, however, is not always indicative of an
inactive nitrate reductase enzyme, suggesting that other mechanisms, such
as exclusion, can yield the resistant phenotype (Murphy and Imbrie, 1981;
Buchanan and Wray, 1982; Marton et al., 1982) (see also Chapter 5 in this
volume).
Wildholm (1981) tried to develop a positive selection system with
tobacco and carrot cells for tryptophan synthase-deficient mutants which
would not convert non-toxic compounds to toxic tryptophan analogues.
However, none of the EMS-treated cells which survived 5-fluoroindole or
6-fluoroindole selection had a tryptophan requirement for growth. The use
here of diploid cell lines may have mitigated against finding recessive
mutations.
When isolating sodium-dependent variants from haploid soybean cell
cultures, Jia-Ping et al. (1981) commented upon the link between enhanced
tolerance and raised requirement for a compound normally needed for
growth but toxic in excess. This is a phenomenon which perhaps could be
exploited more than it has been in auxotroph isolation.
D'Souza and Maher (1976) identified invertaseless mutants in Arabi-
dopsis by spraying aseptically grown seedlings with a sucrose solution fol-
lowed by a special reagent: putative mutants would not produce the red
colour reaction typical of wild-type plants. To date, differential colour
reactions have not been exploited for mutant identification in in vitro work
- potential which perhaps has been underexploited.
C. Negative Selection
i) Total Isolation
It is extremely unfortunate that it has not been possible, so far, to adapt for
the plant cell system the one technique - replica plating (Lederburg and
Lederburg, 1952) - which has so simplified the requirement testing of pu-
tative auxotrophs in microbiology. Although there has been the occasional
report of a replica plating method for plant cells (e. g. Schulte and Zenk,
1977) these have not found widespread applicability. Scientists searching
for auxQtrophs, therefore, have either spent considerable time and effort
trying to develop mutant enrichment techniques or have resorted to the
time and labour intensive task of individual clone testing which has, never-
theless, proved very effective (Savage et al., 1979; Gebhardt et aI., 1981;
Sidorov et al., 1981; Roth and Lark, 1982).
ii) Enrichment
Mutant enrichment is a technique whereby conditions are imposed on a
cell population which preferentially kill growing wild-type cells. The sur-
vivors of the treatment should contain an enhanced proportion of mutants
over the original population thus reducing the subsequent time and labour
spent testing clones.
For both microbiology and animal somatic genetics several enrichment

procedures have been developed. For microorganisms these include sub-
stances probably specific for bacteria and fungi e. g. penicillin (Davis,
1948; Lederburg and Zinder, 1948). Others could have more general appli-
cability, including tritium-labelled thymidine (Lubin, 1959) and nystatin
(Snow, 1966). Ferenczy et al. (1975) developed a method of enriching for
fungal mutants by the lysis of protoplasts selectively produced from
growing colonies. The enrichment achieved was about 80-fold. An unsuc-
cessful attempt was made in our laboratory to apply this procedure to Rosa
cells, but it was not possible to reliably preferentially obtain protoplasts
from dividing cells. Walton et al. (1979) developed an enrichment method
for auxotrophic mutants of yeast based on the fact that stationary phase
yeast are more resistant than exponentially growing cells to killing by heat.
The spectrum of successful "enrichment" compounds was extended by
mammalian cell biologists and includes the nucleoside analogues BUdR
(Puck and Kao, 1967) and FUdR (Basilico and Meiss, 1974). Thymine-less
death of growing cells has also been used (De Mars and Hooper, 1960;
Chu et al., 1969) as a selection procedure.
A further compound, arsenate, was reported to have been successful as
a counter-selective agent enriching for photosynthetic mutants of Euglena
gracilis (Shneyour and Avron, 1975).
Of the agents listed above, nucleoside analogues and arsenate have
proved most popular as counter-selective agents chosen by plant cell biolo-
gists. However, as two recent reviews point out (Blonstein and King, 1985;
Horsch and King, 1985), there is no real evidence that either BUdR or
arsenate have enriched for mutant cells. All too frequently data are not
published for the mutant frequency before treatment by the agent, making
a proper evaluation of the claims for enrichment impossible.
The possible reasons for the failure of several laboratories to establish
reliable and repeatable enrichment techniques, even after several years of
attempts, are numerous. That the processes by which the nucleoside ana-
logues and arsenate preferentially kill growing plant cells have not been
elucidated makes it difficult to identify those components of a system
which might be altered to improve its efficacy. For example, with plant
cells, light exposure is not needed as a post-treatment after BUdR to ensure
killing as it is for animal cells, suggesting that its mode of action in plants
may be different from that established in animal cells (Shillito et al., 1981;
Negrutiu et al., 1985).
Another problem has been that in establishing enrichment conditions it
has usually been necessary to use nitrogen-starved non-growing cells as
models for auxotrophic mutants. Clearly the biochemistry of such cells
may differ substantially from that of auxotrophs, and conditions which
select for the former may be ineffective for the latter. However, now that a
range of tissue-culture expressed auxotrophs are available these can be
used profitably in reconstruction experiments to improve the protocols
although, again, it must not be assumed that anyone auxotroph will be a
good model for all others.
Table 1. Amino-Acid, Vitamin and Purine Auxotrophs Isolated In Vitro. tv
0'1
00
Line Fertile . Spontaneous Key
Species Mutagen Requirement Site of lesion
code pan
I t s GenetIcs' reversion reference(s)
Nicotiana 0.25 % EMS/ Hypoxanthine Yes R? Carlson et al.
tabacum Ih (1973) and
Lysine Yes R? Carlson (1970)
Hyoscyamus MNNG Tryptophan VIIIB9 A reaction of tryptophan No R (in fusion) No Gebhardt et al.
muticus synthetase 3.8 x 10- 8) (1983)
MNNG Histidine VA5 ) imi",",ol"l,"'rolpho,p"" No R (in fusion) No Gebhardt et al.
dehydratase 5.0 x 10- 8) (1983)
XIIIB5 or No R (in fusion)
histidinol-phosphate
IIIE3 aminotransferase No R (in fusion)
MNNG Histidine P03 probably histidinol- No R (in fusion) Yes Shimamoto and ~
phosphate allelic to VA5 (3.2 x 10- 7) King (1983) ~
aminotransferase t;t;
MNNG Leucine XIVE9 No Gebhardt et al. 0'
(also (1982) ::s
en
NR-)
....
('11
MNNG Nicotinamide R (in fusion) Gebhardt et al.

S
I1I2FlO ) before
quinilinic No ) and not and
IVH2 acid No allelic P.I.King
(pers. comm.)
XEI probably quinolinic No R (in fusion)
acid transphosphoribo- and not allelic
sylase to III 2FIO and
IVH2
Glycine max UV Asparagine or Sin-asn- 4 x 10- 7) Roth and Lark
Glutamine (1982)
Datura innoxia 0.5 % EMS/ Isoleucine/ IV-I dihydroxyacid hydratase No Yes Horsch and King
1.5 h valine (1983) and
R. Wallsgrove
(pers. comm.)
Line Fertile G r a Spontaneous Key
Species Mutagen Requirement Site of lesion plants ene ICS reversion reference(s)
code
0.05 % EMS/ Adenine Adl possibly in pathway lead- No 1.3 x 10- 7) King et at. (1980)
Ih ing to 5-aminoimidazole-
4-carboxamide
0.05 % EMS/ Pantothenate Pnl between a-ketoisovalerate No No Savage et al.
Ih and pantoic acid 2.5 x 10- 7) (1979) and
King et al.(l980)
0.25 % EMS/ Threonine JM3 homoserine kinase or Horsch and King
2.5 h threonine synthetase (1984)
Nicotiana UV Isoleucine 6.5/10 Yes R (in fusion) Negrutiu et al.
plumbaginifolia 250 erg cm - 2 6.817 ) threonine deaminase Yes R (in fusion) (1984, 1985) ~
~
4.119 Yes R (in fusion) 0
....
'"i
UV Leucine NI2 as above 0
) 2-0H-3-carboxy- '"0
(as above) isocaproate-isomerase ::r
7.217 or-dehydrogenase
.....
en
0
UV Histidine Ql6 ) histidinol- Yes R (in fusio ll) as above ....~
(as above) 2.1112 phosphate o
3.1114 aminotransferase 2.0 x 10- 7 ::
UV Valine Kl as above
;;-
(as above) plus Ql4 ) valine transaminase ~
.....
Isoleucine 7.3/10 ~
UV Methionine 10.119 ~-cystathionase Yes R (in fusion) as above
(as above)
UV Tryptophan 10.1110 tryptophan synthase Yes R (in fusion) as above
(as above)
60CO y-rays Uracil URA401 No R (in fusion) Sidorovand
Maliga (1982)
60CO y-rays Isoleucine ILE401 L-threonine deaminase No R (in fusion) Sidorov et al.
(1981)
60CO y-rays Leucine LEU403 Yes? R (in fusion) Sidorov und
Maliga (1982)
N
0\
a R = recessive 10
270 A. D. Blonstein
Other factors that have been cited as problematic in enrichment include

density dependent effects (including cross feeding) (Horsch and King,
1984, 1985) and resistance to the enrichment agent. It has been suggested
that BUdR may be poorly incorporated into plant DNA, but that in combi-
nation with FUdR and adenosine or uridine, uptake may be improved
(Zryd, 1976; M. Suter, FMI, pers. comm.). Similarly, lowering phosphate in
the medium may increase the uptake and effectiveness of arsenate (Horsch
and King, 1985).
Mention should be made of the specialised method of enrichment
which has been used successfully by two groups in the isolation of carrot
cell lines showing temperature-sensitive impairment of embryogenesis
(Terzi et ai., 1982; Breton and Sung, 1982). In this regime cells are grown
under embryo-inducing conditions at the chosen restrictive temperature.
Embryos are removed by filtration and the culture shifted to a permissive
temperature. Regenerating embryos are cloned and retested for the temper-
ature-sensitive embryogenic phenotype.
Despite the poor success so far with enrichment schemes for plant cell
mutants, effort still needs to be expended in this field if the search at the
cell level for rare auxotrophic phenotypes is ever to become a routine
procedure.
IV. Phenotypes
A. Amino-Acid, Vitamin and Purine Auxotrophs

Table 1 summarises the amino-acid, vitamin and purine auxotrophs that
have been isolated in the last fifteen years using in vitro techniques. Bearing
in mind that until 1970 only thiamine and proline mutants had been iden-
tified in higher plants, it can be seen that plant tissue culture technology
has been quite successful in increasing the range of mutant and variant
material.
However, these achievements need to be viewed critically. Although the
selection schemes were designed to screen for mutants requiring any of the
20 amino acids, so far lines with confirmed requirements for only 9 have
been isolated - tryptophan, histidine, leucine, isoleucine, valine, thre-
onine, methionine and asparagine and/or glutamine. The possible reasons
for this shortfall have already been discussed. Tissue culture may not
always provide the needs of the auxotroph seeker.
Even more disturbing has been the failure to convincingly demonstrate,
in anyone case, the Mendelian transmission of the variant phenotypes.
Although the methionine- trait in Nicotiana piumbaginifolia was trans-
mitted to all the progeny of selfed regenerants (Negrutiu et ai., 1985) it
failed, as did isoleucine - and histidine-, to segregate in fusion products.
Hyoscyamus muticus nicotinamide- also failed to reappear in the progeny
of a fusion hybrid with NR - N. tabacum (Potrykus et ai., 1984). As anther
culture to rescue auxotrophic meiotic products is unlikely to be acceptable
as standard procedure, it may never be possible to convincingly establish

the genetic basis of these fundamental lesions, let alone map them physi-
cally, and conduct a linkage analysis.
In addition to these complications it can be seen from Table 1 that, with
the exception of Carlson's possible tobacco variants, all the amino-acid
mutations have been induced in a narrow range of so-called "model"
species which, although amenable to tissue culture, have no prior estab-
lished genetics or, of lesser import, economic value. The rather miserable
spectre arises that when species which fulfil these conditions (e. g. tomato)
have been sufficiently modified for tissue culturability, the whole mutant
induction and screening procedure will take place all over again.
So far, the amino-acid auxotrophs have been isolated by tissue culture
scientists intent on demonstrating that in vitro techniques can produce vari-
ation that had not yet been obtained at the whole plant level. They have
also aimed to provide tissue-culture markers which would assist in such
techniques as fusion and transformation. The mutant isolators have not
been the biochemists intimately concerned with specific areas of plant
metabolism. They would not only benefit most directly from the avail-
ablility of genetic variation within their field of study, but might also, from
more profound chemical understanding, be better equipped to design
effective, specific selection systems, especially positive schemes which
could take advantage of the vast array of metabolite analogues that can be
produced synthetically.
B. Nitrate Reductase Mutants

As discussed in Section III. B, great success has been achieved in the iso-
lation of nitrate reductase-deficient auxotrophs by taking advantage of the
positive chlorate resistance selection scheme. These mutants, and the role
they have played in the elucidation of nitrogen metabolism in plants are
fully discussed by John Wray in Chapter 5 of this volume and will not be
treated further here.
C. Temperature-Sensitive Mutations
Temperature-sensitive (ts) mutations have contributed significantly to the
understanding of many aspects of the regulation of development from
micro-organisms (e. g. Edgar and Lielausis, 1964) to Drosophila (Suzuki et
at., 1976). Ts lines have also been obtained in mammalian somatic genetics
(Naha, 1969; Thompson et at., 1970) and a few in plant cell work, which
are described below.
Two advantages accrue to ts mutations. First, they facilitate the reso-
lution of the temporal activation of a series of related genes and second,
the expression of a potentially very disruptive lesion can be suppressed
while regeneration, crosses etc. are carried out. Bearing in mind the
problems with regeneration and genetic analysis of the auxotrophic muta-
tions that have been described in the previous section, the advantages of ts
272 A. D. Blanstein
conditional lethals should be quite clear. However, even with an effective

mutagenesis procedure one can expect ts auxotrophs to be quite rare
emphasising further the urgent need for reliable enrichment procedures.
The most effective use to date of in vitro isolated temperature-sensitive
mutants have been the studies by the groups of Z. R. Sung at Berkeley and
M. Terzi in Pisa isolating ts mutants in carrot cell embryogenesis, which are
helping them build up a time-sequence of the molecular and biochemical
events which accompany and drive embryogenesis (Breton and Sung, 1982;
Terzi et al., 1982; Giuliana et a!., 1984). The enrichment by filtration
procedure they used has already been described (see Ill. C. ii). Because
these mutants are not strictly auxotrophs they will not be discussed in any
detail here, although- one (ts 66) isolated by Breton and Sung (1982), but
subsequently lost, may have been an auxin auxotroph sinc.e it was unable
to grow at all in the absence of 2,4-D in contrast to the wild type and other
ts-emb- lines.
A ts auxin auxotroph of Hyoscyamus muticus, XIIB2, which, unfortu-
nately, has not yet been regenerated to fertile plants, is being studied in our
laboratory with the aim of elucidating certain aspects of auxin production
and action in plant cells. XIIB2 was originally isolated as an uncharac-
terised ts line in a large auxotroph isolation experiment (Gebhardt et al.,
1981). It is recessive in fusion, and plantlets that have been produced also
exhibit the ts phenotype. It was subsequently found that the callus' phe-
notypic expression of browning followed by death 4 to 5 days after the
temperature shift to 33 C could be corrected by the exogenous addition of
a range of auxins, including indole-3-acetic acid. (The wild type is
hormone autotrophic in culture.) It does not, however, appear to be a
simple biosynthetic lesion because RIA has detected normal levels of IAA
in XIIB2 at the high temperature (1. Oetiker, pers. comm.). We hope to
produce more such ts auxotrophs to facilitate studies of hormone biosyn-
thesis and action.
Malmberg (1979, 1980) also produced and partially characterized some
in vitro derived ts mutants of Nicotiana tabacum, but this work has
remained incomplete.
Temperature-sensitivity need not be the only approach to manipulating
the expression of conditionallethals such as auxotrophs. Other factors that
might be considered, for example, are pH and osmotic pressure.
v. Fusion and Transformation

From the start, attempts were not made to produce auxotrophs in culture
solely to provide material for chemical analysis. It was also hoped that
stable auxotrophic cell lines with low (preferably absent) reversion rates
could be used as models in the refinement of fusion and transformation
techniques.
So far, a great deal more has been done with fusion than with transfor-
mation, and for recent reviews on the technology, somatic hybridization,
and segregation of organelles and cytoplasmic traits in fusion products see

Ulzar (1983), Harms (1983) and Fluhr (1983).
One use of fusion can be to attempt to regenerate a line which will not
do so otherwise, because it has been found that fusion hybrids can often
regenerate when only one of the two parents is so capable (Evans, 1983;
Potrykus et al., 1984).
Probably the most extensive fusion work has been done with the NR-
lines (see Chapter 5 in this volume). Where plant regeneration is not pos-
sible, fusion will provide the only means of determining dominance rela-
tionships and allele complementation (see Table 1).
Although for transformation non-revertant auxotrophs should have
great advantages over leaky resistance phenotypes (Shillito et al., 1983),
there have been no confirmed reports yet of auxotroph complementation
by transformation using genes of either microbial or plant origin.
VI. Conclusions and Future Prospects
In vitro culture techniques with cells have undoubtedly produced new plant
variation which it might have been very difficult, if not impossible, to
isolate at the whole plant level. In some, but not all, cases the new material
has played an important role in the elucidation of certain biochemical pro-
cesses in plants, and further isolation of auxotrophs seems justified if a
number of criteria are met.
Auxotroph isolation should be "targeted". That is, cell populations
should be screened for specific phenotypes intended as tools to study given
processes.
In the absence of positive selection, enrichment procedures are
essential to reduce effort and frustration, and to maximise the range of
variation that can be isolated.
To ensure that a system can be properly analysed genetically it seems
very advisable to work with variation that has additional conditionality
(e. g. temperature-sensitivity) so that material can be handled with relative
ease at the whole plant level.
Given these three objectives, targeting, enrichment and a conditionality
additional to the auxotrophy, in vitro isolation of auxotrophs continues to
hold promise as a technique contributing "Genetics" to the study of "Bio-
chemistry" .
Acknowledgements
I would like to thank Dr. Patrick J. King and other colleagues at the FMI
for helpful discussion in the preparation of this chapter, and Dr. John King
(Saskatoon) for providing me with unpublished results.
274 A. D. Blonstein
VII. References
Aberg, B., 1947: On the mechanism of the toxic action of chlorates and some
related substances upon young wheat plants. Ann. R Agric. ColI. Sweden 15,
37-107.
Basilico, C., Meiss, H. K., 1974: Methods for selecting and studying temperature-
sensitive mutants of BHK-21 cells. In: Prescott, D. M. (ed.), Methods of Cell
Biology, Vol, 8, pp. 1-22. New York: Academic Press.
Behnke, M., 1976: Kulturen isolierte Zellen von einigen dihaploiden Solanum-
tuberosum-KIonen und ihre Regeneration. Z. Pflanzenphysiol. 78, 177 -181.
Binding, H., 1974: Regeneration von haploiden und diploiden Pflanzen aus Proto-
plasten von Petunia hybrida. Z. Pflanzenphysiol. 74, 327 -356.
Binding, H., Nehls, R, Schieder, 0., Sopory, S. K., Wenzel, G., 1978: Regeneration
of mesophyll protoplasts isolated from dihaploid clones of Solanum tuberosum.
Physiol. Plant. 43, 52-54.
Blonstein, A. D., King, P. J., 1985: Plant mutant isolation via protoplasts. In:
Fowke, L. c., Constabel, F. (eds.), Plant Protoplasts, pp. 151-168. Boca Raton,
Flo.: CRC Press.
Bourgin, J.-P., Chupeau, Y., Missonier, C., 1979: Plant regeneration from meso-
phyll protoplasts of several Nicotiana species. Physiol. Plant. 45, 288-292.
Breton, A. M., Sung, Z. R, 1982: Temperature-sensitive carrot variants impaired in
somatic embryogenesis. Dev. BioI. 90, 58-66.
van Breukelen, E. W. M., Ramanna, M. S., Hermsen, J. G. Th., 1975: Mono-
haploids (n=x= 12) from autotetraploid Solanum tuberosum (2n=4x=48)
through two successive cycles of female parthenogenesis. Euphytica 24,
567-574.
Buchanan, R J., Wray, J. L., 1982: Isolation of molybdenum cofactor defective cell
lines of Nicotiana tabacum. Mol. Gen. Genet. 188,228-234.
Carlson, P. S., 1970: Induction and isolation of auxotrophic mutants in somatic cell
cultures of Nicotiana tabacum. Science 168,487-489.
Carlson, P. S., Dearing, R. D., Floyd, B. M., 1973: Defined mutants in higher
plants. In: Srb, A. M. (ed.), Genes, Enzymes and Populations, pp. 99-107.
New York: Plenum.
Chu, E. H. Y., Brimer, P., Jacobsen, K. B., Merriam, E. V., 1969: Mammalian cell
genetics I. Selection and characterization of mutations auxotrophic for L-glu-
tamine or resistant to 8-azaguanidine in Chinese hamster cells in vitro. Genetics
62, 359-377.
Colijn, C. M., Kool, A. J., Nijkamp, H. J. J., 1979: An effective chemical muta-
genesis procedure for Petunia hybrida cell suspension cultures. Theor. Appl.
Genet. 55, 10 1-106.
Davis, B. D., 1948: Isolation of biochemically deficient mutants of bacteria by peni-
cillin. J. Am. Chem. Soc. 70, 4267.
De Mars, R., Hooper, J. L., 1960: A method of selecting for auxotrophic mutants of
HeLa cells. J. Exp. Med. 111, 559-573.
D'Souza, E. S., Maher, E. P., 1976: An invertase mutant of Arabidopsis. Arabidopsis
Inf. Servo 13, 79-84.
Dunwell, J. M., Sunderland, N., 1973: Anther culture of Solanum tuberosum L.
Euphytica 22, 317-323.
Durand, J., 1979: High and reproductive plating efficiencies'of protoplasts isolated
from in vitro grown haploid Nicotiana sylvestris Spegaz et Comes. Z. Pflanzen-
physiol. 93, 283-295.
Durand, J., Potrykus, I., Donn, G., 1973: Plantes issues de protoplastes de Petunia.
Z. Pflanzenphysiol. 69, 26-34.
Edgar, R. S., Lielausis, I., 1964: Temperature sensitive mutations of bacteriophage
T4D: Their isolation and genetic characterization. Genetics 49, 649-662.
Engel, P. P., 1968: The induction of biochemical and morphological mutants in the
moss Physcomitrella patens. Am. J. Bot. 55, 438-446.
Evans, D. A., 1983: Agricultural applications of plant protoplast fusion. Bio/Tech-
nology 1,253-261.
Feenstra, W. J., 1964: Isolation of nutritional mutants in Arabidopsis thaliana. Ge-
netica 35, 259-269.
Ferenczy, L., Sipiczki, M., Szegedi, M., 1975: Enrichment of fungal mutants by
selective cell-wall lysis. Nature (London) 253, 46-47.
Fluhr, R., 1983: The segregation of organelles and cytoplasmic traits in higher plant
somatic fusion hybrids. In: Potrykus, I., Harms, C. T., Hinnen, A., Hutter, R.,
King, P. J., Shillito, R. D. (eds.), Protoplasts 1983. Lecture Proceedings (6th Inter-
national Protoplast Symposium, Basel 1983), pp. 85-92. Basel: Birkhauser.
Furner, I. J., King, J., Gamborg, O. L., 1978: Plant regeneration from protoplasts
isolated from a predominantly haploid suspension culture of Datura innoxia.
Plant Sci. Lett. 11, 169-176.
Galbraith, D. W., Harkins, K. R., Maddox, J. M., Ayres, N. M., Sharma, D. P.,
Firoozabady, E., 1983: Rapid flow cytometric analysis of the cell cycle in intact
plant tissues. Science 220, 1049-1051.
Galun, E., Raveh, D., 1975: In vitro culture of tobacco protoplasts: Survival of
haploid and diploid protoplasts exposed to X-ray radiation at different times
after isolation. Radiat. Bot. 15, 79-82.
Gavazzi, G., Nava-Racchi, M., Tonelli, C., 1975: A mutation causing proline
requirement in Zea mays. Theor. Appl. Genet. 46, 339-345.
Gebhardt, c., Schnebli, V., King, P. J., 1981: Isolation of biochemical mutants
using haploid mesophyll protoplasts of Hyoscyamus muticus. II. Auxotrophic
and temperature-sensitive clones. Planta 153, 81-89.
Gebhardt, C., Fankhauser, H., King, P. J., 1982: Amino-acid auxotrophs of Hyos-
cyamus muticus and their characterisation. In: Fujiwara, A. (ed.), Plant Tissue
Culture 1982. Proceedings of the 5th International Congress of Plant Tissue
and Cell Culture, Tokyo, 1982, pp. 463-464. Tokyo: Maruzen Co.
Gebhardt, c., Shimamoto, K., Lazar, G., Schnebli, V., King, P. J., 1983: Isolation of
biochemical mutants using haploid mesophyll protoplasts of Hyoscyamus
muticus III. General characterisation of histidine and tryptophan auxotrophs.
Planta 159, 18-24.
Giuliano, G., Lo Schiavo, F., Terzi, M., 1984: Isolation and developmental charac-
terization of temperature-sensitive carrot cell variants. Theor. Appl. Genet. 67,
179-183.
Gowans, C. S., 1960: Some genetic investigations on Chlamydomonas eugametos. Z.
Vererbungsl. 91, 63-73.
Gresshof, P. M., 1978: Auxotrophic mutant isolation in higher plants - a thought
experiment with Arabidopsis thaliana. Arabidopsis Inf. Servo 15, 88-91.
Haas, F., Mitchell, M. B., Ames, B. N., Mitchell, H. K., 1952: A series of histi-
dineless mutants of Neurospora crassa. Genetics 37, 217-226.
Harms, C. T., 1983: Somatic hybridization by plant protoplast fusion. In: Potrykus,
I., Harms, C. T., Hinnen, A., Hutter, R., King, P. J., Shillito, R. D. (eds.), Proto-
plasts 1983. Lecture Proceedings (6th International Protoplast Symposium,
Basel 1983), pp. 69-84. Basel: Birkhauser.
276 A. D. Blonstein
Horsch, R. B., King, J., 1983: Isolation of an isoleucine-valine-requiring auxotroph

from Datura innoxia cell cultures by arsenate counterselection. Planta 159,
12-17.
Horsch, R. B., King, J., 1984: The isolation of auxotrophs from Datura innoxia Mill.
cell cultures following recovery of arsenate-treated cells on feeder plates. Planta
160, 168-173.
Horsch, R. B., King, J., 1985: Arsenate counters elective enrichment for auxotrophic
plant cells works well in theory but not in practice. Can. J. Bot. 63, 2115-2120.
Jia-Ping, Z., Roth, E. J., Terzaghi, W., Lark, K. G., 1981: Isolation of sodium
dependent variants from haploid soybean cell culture. Plant Cell Rep. 1,
48-51.
Kao, K. N., Michayluk, M. R., 1975: Nutritional requirements for growth of Vicia
hajastana cells and protoplasts at very low population density in liquid media.
Planta 126, 105-110.
King, J., Khanna, V., 1980: A nitrate reductase-less variant isolated from
suspension cultures of Datura innoxia (Mill.). Plant Physiol. 66,632-636.
King, J., Horsch, R. B., Savage, A. D., 1980: Partial characterization of two stable
auxotrophic cell strains of Datura innoxia Mill. Planta 149, 480-484.
Lamm, R., 1938: Notes on a haploid potato hybrid. Hereditas 24, 391-396.
Langridge, J., Brock, R. D., 1961: A thiamine-requiring mutant of the tomato. Aust.
J. BioI. Sci. 14, 66-69.
Lftziu, G. B., 1983: Recent developments in plant protoplast fusion and selection
technology. In: Potrykus, I., Harms, C. T., Hinnen, A., Hutter, R., King, P. J.,
Shillito, R. D. (eds.), Protoplasts 1983. Lecture Proceedings (6th International
Protoplast Symposium, Basel 1983), pp. 61-67. Basel: Birkhauser.
Lederburg, J., Lederburg, E. M., 1952: Replica plating and indirect selection ofbac-
terial mutants. J. Bacteriol. 63, 399-406.
Lederburg, J., Zinder, N., 1948: Concentration of biochemical mutants of bacteria
with penicillin. J. Am. Chern. Soc. 70, 4267 -4268.
Lein, J., Mitchell, H. K., Houlahan, M. B., 1948: A method for selection of bio-
chemical mutants of Neurospora. Genetics 34, 435-442.
Li, S. L., Redei, G. P., 1969: Thiamine mutants of the crucifer, Arabidopsis.
Biochem. Genet. 3, 163-170.
Li, S. L., Redei, G. P., Gowans, C. S., 1967: A phylogenetic comparison of mutation
spectra. Mol. Gen. Genet. 100,77-83.
Lorz, H., Wernicke, W., Potrykus, I., 1979: Culture and plant regeneration of Hyos-
cyamus protoplasts. Planta Med. 36, 21-29.
Lubin, M., 1959: Selection of auxotrophic bacterial mutants by tritium-labeled
thymidine. Science 129, 838-839.
Magnien, E., Devreux, M., 1980: A critical assessment of the protoplast system as a
tool for radiosensitivity studies. In: Sala, F., Parisi, B., Cella, R., Ciferri, O.
(eds.), Plant Cell Cultures: Results and Perspectives, pp. 121-126. Amsterdam:
Elsevier.
Magnien, E., Dalschaert, X., Devreux, M., 1980: Different radiosensitivities of
Nicotiana plumbaginifolia leaves and regenerating protoplasts. Plant Sci. Lett.
19,231-241.
Malmberg, R. L., 1979: Temperature-sensitive variants of Nicotiana tabacum iso-
lated from somatic cell culture. Genetics 92, 215-221.
Malmberg, R. L., 1980: Biochemical, cellular and developmental characterization
of a temperature-sensitive mutant of Nicotiana tabacum and its second site
revertant. Cell 22, 603-609.
Marton, L., Dung, T. M., Mendel, R. R., Maliga, P., 1982: Nitrate reductase defi-
cient cell lines from haploid protoplast cultures of Nicotiana plumbaginifolia.
I Mol. Gen. Genet. 182,301-304.
Muller, A. J., Grafe, R., 1978: Isolation and characterization of cell lines of Nico-
tiana tabacum lacking nitrate reductase. Mol. Gen. Genet. 161, 67-76.
Murphy, T. M., Imbrie, C. W., 1981: Induction and characterization of chlorate-
resistant strains of Rosa damascena cultured cells. Plant Physiol. 67, 910-916.
Murr, S. M., Spurr, A. R., 1973: An albino mutant in Plantago insularis requiring
thiamine pyrophosphate. J. Exp. Bot. 24, 1271-1282.
Murr, S. M., Stebbins, G. L., 1971: An albino mutant in Plantago insularis requiring
thiamine pyrophosphate. I: Genetics. Genetics 68, 231-243.
Nagata, T., Takebe, I., 1971: Plating of isolated tobacco mesophyll protoplasts on
agar medium. Planta 99, 12-20.
Nagy, J. I., Maliga, P., 1976: Callus induction and plant regeneration from meso-
phyll protoplasts of Nicotiana sylvestris. Z. pflanzenphysiol. 78, 453-455.
Naha, P. M., 1969: Temperature sensitive conditional mutants of monkey kidney
cells. Nature (London) 223, 1380-1381.
Nasim, A., Hannan, M. A., Nestmann, E. R., 1981: Pure and mosaic clones - a re-
flection of differences in mechanisms of mutagenesis by different agents in Sac-
charomyces cerevisiae. Can. J. Genet. Cytol. 23,73-79.
Negrutiu, I., 1981: Improved conditions for large-scale culture, mutagenesis and
selection of haploid protoplasts of Nicotiana plumbaginifolia Viviani. Z. pflan-
zenphysiol. 104, 431-442.
Negrutiu, I., Dirks, R., Jacobs, M., 1983: Regeneration of fully nitrate
reductase-deficient mutants from protoplast culture of Nicotiana plumbagin i-
folia (Viviani). Theor. Appl. Genet. 66, 341-347.
Negrutiu, I., Jacobs, M., Cattoir-Reynaerts, A., 1984: Progress in cellular
engineering of plants: biochemical and genetic assessment of selectable
markers from cultured cells. Plant Mol. BioI. 3, 289-302.
Negrutiu, I., de Brouwer, D., Dirks, R., Jacobs, M., 1985: Amino acid auxotrophs
from protoplast cultures of Nicotiana plumbaginifolia Viviani. I. BUdR
enrichment selection, plant regeneration, and general characterisation. Mol.
Gen. Genet. 199,330-337.
Nelson, O. E., Burr, B., 1973: Biochemical genetics of higher plants. Annu. Rev.
Potrykus, I., Jia, J., Lazar, G. B., Saul, M., 1984: Hyoscyamus muticus + Nicotiana
tabacum fusion hybrids selected via auxotroph complementation. Plant Cell
Rep.3,68-71.
Puck, T. T., Kao, F.-T., 1967: Genetics of somatic mammalian cells, V. Treatment
with 5-bromodeoxyuridine and visible light for isolation of nutritionally defi-
cient mutants. Proc. Nat. Acad. Sci., U.S.A. 58,1227-1234.
Racchi, M. L., Gavazzi, G., Monti, D., Manitto, P., 1978: An analysis of the nutri-
tional requirements of the pro mutant in Zea mays. Plant Sci. Lett. 13, 357 -364.
Raveh, D., Hubermann, E., Galun, E., 1973: In vitro culture of tobacco protoplasts:
use of feeder techniques to support division of cells plated at low densities. In
Vitro 9, 216-222.
Redei, G. P., 1965: Genetic blocks in the thiamine synthesis of the angiosperm
Arabidopsis. Am. J. Bot. 52, 834-841.
Redei, G. P., 1975: Induction of auxotrophic mutants in plants. In: Ledoux, L.
(ed.), Genetic Manipulations with Plant Material, pp. 329-350. New York:
Plenum.
278 A D. Blonstein
Roth, E. J., Lark, KG., 1982: Isolation of an auxotrophic cell line of soybean
(Glycine max) which requires asparagine or glutamine for growth. Plant Cell
Rep. 1, 157-160.
Savage, A D., King, J., Gamborg, O. L., 1979: Recovery of a pantothenate auxo-
troph from a cell suspension culture of Datura innoxia Mill. Plant Sci. Lett. 16,
367-376.
Schieder, 0., 1975: Regeneration von haploiden und diploiden Datura innoxia Mill.
Mesophyll-Protoplasten zu Pflanzen. Z. Pflanzenphysiol. 76, 462-466.
Schieder, 0., 1976: The spectrum of auxotrophic mutants from the liverwort
Sphaerocarpos donnelli Aust. Mol. Gen. Genet. 144,63-66.
Schulte, U., Zenk, M. H., 1977: A replica plating method for plant cells. Physiol.
Plant. 39, 139-142.
Shahin, E. A., 1985: Totipotency of tomato protoplasts. Theor. Appl. Genet. 69,
235-240.
Shepard, J. F., 1980: Mutant selection and plant regeneration from potato meso-
phyll protoplasts. In: Rubenstein, I., Gengenbach, B., Phillips, R. L., Green, C.
E. (eds.), Genetic Improvement of Crops. Emergent Techniques, pp. 185-219.
Minneapolis: Univ. of Minnesota Press.
Shields, R., 1980: Getting out of the wood? Cell 19,303-304.
Shillito, R. D., Street, H. E., Schilperoort, R. A, 1981: Model system studies of the
use of 5-bromo-2' -deoxyuridine for selection of deficient mutants in plant cell
suspension and protoplast cultures. Mutat. Res. 81, 165-175.
Shillito, R. D., Ulzar, G., Paszkowski, J., Shimamoto, K., Nicola-Koukolikova, Z.,
Hohn, B., Hohn, T., Potrykus, I., 1983: Approaches to plant protoplast transfor-
mation using drug resistance and auxotroph complementation as selective
markers. In: Lurquin, P. F., Kleinhofs, A (eds.), Genetic Engineering in Euka-
ryotes, pp. 265-276. New York: Plenum.
Shimamoto, K, King, P. J., 1983: Isolation of a histidine auxotroph of Hyoscyamus
muticus during attempts to apply BUdR enrichment. Mol. Gen. Genet. 189,
69-72.
Shneyour, A, Avron, M., 1975: A method for producing, selecting, and isolating
photosynthetic mutants of Euglena gracilis. Plant Physiol. 55, 142-144.
Shneyour, Y., Zelcer, A, Izhar, S., Beckmann, J. S., 1984: A simple feeder-layer
technique for the plating of plant cells and protoplasts at low density. Plant Sci.
Lett. 33, 293-302.
Sidorov, V. A, Maliga, P., 1982: Fusion-complementation analysis of auxotrophic
and chlorophyll-deficient lines isolated in haploid Nicotiana plumbaginifolia
protoplast cultures. Mol. Gen. Genet. 186, 328-332.
Sidorov, V., Menczel, L., Maliga, P., 1981: Isoleucine-requiring Nicotiana plant
deficient in threonine deaminase. Nature (London) 294, 87 -88.
Snow, R., 1966: An enrichment method for auxotrophic yeast mutants using the
antibiotic "Nystatin". Nature (London) 211, 206-207.
Sopory, S. K, Jacobsen, E., Wenzel, G., 1978: Production of mono haploid embryos
and plantlets in cultured anthers of Solanum tuberosum. Plant Sci. Lett. 12,
47-54.
Suzuki, D. T., Kaufman, T., Falk, D., and the U. B. C. Drosophila Research Group,
1976: 5. Conditionally expressed mutations in Drosophila melanogaster. In:
Ashburner, M., Novitski, E. (eds.), The Genetics and Biology of Drosophila,
Vol. lA, pp. 207-263. London: Academic Press.
Terzi, M., Giuliano, G., Lo Schiavo, F., Nuti Ronchi, V., 1982: Studies on plant cell
lines showing temperature sensitive embryogenesis. In: Burger, M. (ed.),
Embryonic Development, Part B: Cellular Aspects, pp. 521-534. New York:

R. Alan Liss.
Thomas, E., 1981: Plant regeneration from shoot culture-derived protoplasts of
tetraploid potato (Solanum tuberosum cv. Maris Bard). Plant Sci. Lett. 23,
81-88.
Thompson, L. H., Mankovitz, R., Baker, R. M., Till, J. E., Siminovitch, L.,
Whitmore, G. F., 1970: Isolation of temperature-sensitive mutants of L-cells.
Proc. Nat. Acad. Sci., U.S.A. 66, 377-384.
Walton, E. F., Carter, B. L. A., Pringle, J. R., 1979: An enrichment method for tem-
perature-sensitive and auxotrophic mutants of yeast. Mol. Gen. Genet. 171,
111-114.
Weber, G., Lark, K. G., 1979: An efficient plating system for rapid isolation of
mutants from plant cell suspensions. Theor. Appl. Genet. 55, 81-86.
Widholm, J. M., 1981: Utilization of indole analogs by carrot and tobacco cell
tryptophan synthase in vivo and in vitro. Plant Physiol. 67, 1101-1104.
Zryd, J.-P., 1976: 5-bromodeoxyuridine as an agent in the selection of sycamore cell
cultures. Plant Sci. Lett. 6, 157 -161.
Subject Index
ABA - see abscisic acid Agrobaeterium-plant interactions 195
abscisic acid Agrobaeterium rhizogenes 194
abscission 35 Agrobaeterium tumefaciens
accumulation 48, 49 crown gall disease 234
dormancy 35, 39 mutagenesis of plant genes 134
drought stress 36, 38, 48 nodulin induction 170
dwarfness 49 T-DNA234
elongation 35 Ti plasmid 137, 140
geotropism 35 Adh transformation 95
receptor 47 AL-toxin 240
seed content 44-46 albumin 210
sensitivity 46-48
alfalfa - see Medieago sativa
stomatal opening 35
allyl alcohol pollen selection 75, 254
synthesis 36
Alternaria alternata 240
water relations 39, 44-46
Amaranthus hybridus 62
abscisic acid mutants amino acid auxotrophs 270
deficient 38 ff. amylase 47, 220
dormancy 44, 45 Anabaena 67
insensitive 46-48 Anaeystis nidulans 136
precocious germination 44 anaerobic treatment
resistant 47 alcohol dehydrogenase expression
uptake 46 76
abscission 35 transition polypeptides 76, 78
Activator (Ae) 88 apical dominance mutants 49
adenine auxotrophs 268, 269 apical senescence 2, 18
alcohol dehydrogenase 73 ff., 248 apple - see Malus domestiea
(see also Adh) Arabidopsis thaliana
anaerobic expression 76 ABA mutants 38-40, 44
activity 252 ABA receptor 47
developing endosperm 75 ABA-insensitive mutants 47, 48
functional domains 85 AdhJ cloning 82
isozymes of Zea mays 74 auxotrophs 260
organ specificity 75 chlorate resistance 109, 183
pollen 75 dicarboxylate transport mutant 65
mutants 76, 254 dwarf mutants 14
Adh genes:
cloning 78-82, 84 abaJ/3/440-42,45-47
gene regulation 89 ff. abi-J/-2/-3 47
gene structure 82-84 Adh 84, 85,90-92
sequence divergence 82 ehll/2/3 110, 113, 122, 123
ADH - see alcohol dehydrogenase enx 112, 130
Aegi/ops, glutelin genes 218 ga-l/-2/-3/-4/-5 3,4, 17,39,40
Aesehynomene stem nodulation 161 rgn 112, 130
282 Subject Index
gibberellic acid mutants 3, 40 Bradyrhizobium 161, 187

invertaseless mutants 266 auxotrophs 167
nitrate assimilation mutants Il2, 122, B.japonicum 182
129 Glycine max 163
photorespiration mutants 65 mutants 167, 175, 185
seed coat mutants 41 nodule induction 184
seed dormancy 44, 45 nodule types 168
seeds, abscisic acid 45, 46 Parasponia rigida 167
suppressor mutant 39 symbiotic genetics 172, 181
trans-3-hexadecenoic acid deficient Vigna sinensis 175
mutant 62 bread wheat 216, 217, 222
Arachis hypogaea bromo-deoxyuridine
Bradyrhizobium test 175 auxotroph enrichment 267, 270
endosperm protein 208 bud dormancy,35
non-nodulation mutants 178 BUdR - see bromo-deoxyuridine
arsenate, enrichment compound 267
asparagine auxotrophs 268, 269
Aspergillus nidulans Capsicum annuum 40
nitrate assimilation mutants 42, 109, carotenoid biosynthesis 43, 63
110,119,123,127,131-133,140 cell-culture systems 261
nitrate assimilation 102 Centrosema pubescens 176
nitrate reductase 104, 111, 132 Ceratopteris 47
transformation systems 137 chickpea - see Cicer arietinum
ATP synthase 63, 66 Chlamydomonas
atrazine 62 auxotrophs 260
attachment of bacteria 171 bicarbonate transport 64
aUXIn triazine resistance 63
auxotroph 272 chlorate
nodules 191 nitrate reductase mutants 109,265
auxotrophic mutants 17, 166, 259 ff., resistance 109, 133, 181,266
268,269,272 Chlorella
Avena nuda chlorate 109
viviparous mutants 40 nitrate reductase 105
A vena sativa chlorophyll
abnormal karyotypes 223 alb binding proteins 61
avenin 210 protein-1 deficient mutant 64
chromosome mutants 225 chloroplast
endosperm proteins 207, 213 electron transport processes 56
HV-toxin 240 genome 59-61, 67, 68
nitrite reductase 108 homology with mitochondrial DNA
Pc-2 240,241 69
resistance to crown rust 240 mRNA67
storage protein solubility 211 nuclear encoded genes 61
Victoria blight 240, 241 ribosome mutants 63,66
Avenin 210 Cicer arietinum
avirulence genes 235 genes:
jix- 179
bacteriods rnl/4/5 179
Bradyrhizobium 169 leghemoglobin mutants 179
development 165 nitrate sensitivity 175
isolated 169 non-fixation mutants 179
barley - see Hordeum vulgare Citrullus lanatus 4
Subject Index 283
Cochliobolus enrichment 266

C. carbonum, He-toxin 239, 240 auxotrophs 265
C. heterostrophus, T-toxin sensitivity compounds 267
241 Erythrina senegalensis 103, 141
C. miyabeanus 242 ethyl methane sulphonate 76, 117, 182,
C. victoriae 240 264,268,269
toxin production 241 Euglena gracilis
controlling elements 88, 92 arsenate enrichment 267
cotton - see Gossypium chloroplast genome 67
cowpea - see Vigna sinensis exopolysaccharide (EPS) 173
crown gall disease 234
Cucurbita farnesyl pyrophosphate 36
nitrate reductase 104, 106 fixation thread 169
nitrite reductase 108 flax - see Linum usitatissimum
Cyanophora paradoxa, cyanelle DNA flax rust 235
68 flowering
cytokinin, synthesis in Rhizobium 191 gibberellins 18
Pisum sativum 18
Datura innoxia fluoro-deoxyuridine 267, 270
auxotrophs 263, 268, 269 fluridone 41
nitrate reductase mutants 117, 265 Fragaria vesca 5
somatic hybridisation 135 Frankia 169
Daucus carotus fructose-1,6-diphosphate aldolase 78
abscisic acid 46 FUdR - see fluoro-deoxyuridine
auxotroph enrichment 270 Fusarium oxysporum 242
mutants in embryogenesis 272
somatic embryo development 46 gametophytic gene expression 247
Dianthus chinensis 249 gamma ray irradiation 182,263,268,
diaphorase function 107 269
dicarboxylic acid transport 167 gene tagging 238
dihydrophaseic acid (DPA) 36 cloning NR gene 136
Dissociator (Ds) 88, 92 transformation 235
diuron 62 gene-for-gene resistance 235 ff.
Dolichos 175 geotropism 35
dormancy Gibberella fujikuroi 1
abscisic acid mutants 44, 45 gibberellins 1 ff.
mutant selection 39 bioassays 5
seed 2 biologically active 22
Drechslera - see Cochliobolus biosynthesis 8-14, 22, 23
drought stress 48 dormancy 17
dwarf habit 2, 3 dwarf mutants 2 ff.
abscisic acid 49 flowering 18-21
mutants, light and dark influence 24 function in seeds 23
genes influencing sensitivity 3
electron transport 58 internode length 25
embryogenesis mutants 270 photoperiod control 19,21
EMS - see ethyl methane sulphonate receptor sites 16
endosperm proteins 207 ff. gibberellin mutants
development 208 breakdown 6
fractionation 212, 213 insensitive 6, 7,14-16,24
gene regulation 214 tissue specificity 24
genetic variation 211 ff. sensitive 5, 10
284 Subject Index
sites of action 7, 22 pollen competition 248

overproducer 14
gliadin 210, 217 haploid protoplasts 119,261
globulin 210 He-toxin 240
glucose phosphate isomerase 78 heat shock response 76
glutamate oxaloacetate transaminase heavy metal tolerance
microspore selection 253
250
Helianthus annuus 39
glutamine auxotroph 268, 269
Helminthosporium
glutamine synthase 168 H. carbonum 240
cDNA cloning 170 H. oryzae 242
glutelin 210 H. victoriae 240
glutenin 210, 217 heme
Glycine max 161, 181, 187 biosynthesis, Rhizobium mutants 166
abscisic acid sensitivity 46 content, Glycine max 166
anaerobic expression 76, 78 high-lysine lines 215,219,220
autoregulation of nodulation 190 histidine auxotrophs 268, 269
auxotrophs 268, 269 hordein genes 210, 225,226
Bradyrhizobium nodules 174 Hordeum chilense
chlorate sensitivity 109 glutelin genes 218
endosperm protein 208 Hordeum vulgare
genes: abscisic acid sensitivity 47
fix- 182 Adh genes 84, 90, 91
nod- 182, 184 Adh genomic clone 82
nod3 180, 190 a-amylase 220
nrl 113, 122, 131, 132, 183 dormancy 41
nts 182,185-189 genes:
11/2/3/4/5 177,178,184 albostrians 42
grafting and supernodulation 189, br 4, 10, 11
191 chlorina fl 61, 64
leghemoglobin 166, 167 GA-ins 4
mutagenesis 178, 182 GA-less 4, 11
nitrate assimilation mutants 113, 121 gigas 4, 11, 14
nitrate sensitivity 175 Horl/2/2a/2ca/3 218-220,225
nitrate-tolerant mutants 182, lys 220
185-189,194 Ml-a 226
nitrogen fixation 142, 175 narl/2/3/4 113, 121, 122, 128 ff.
nodulation 161-168, 171, 177, 180, sIn 4
181, 190 uz 4, 11
nodulins 170 viridis-b63/-n34 64
oxygen diffusion barrier 165 Xno18/19 128
pseudoinfections 173 gibberellin insensitive mutants 4, 14,
resistance to Pseudomonas syringae 16,24
236 gibberellins in developing seeds 11
seed lectin 185 high-lysine 219
sodium-dependent variants 266 high-protein genes 216
supernodulation 186, 188 hordein 207, 210, 219
symbiosis 165 hordein genes 225, 226
ureide production 168 nitrate reductase 104, 107
Gossypium nitrate reductase mutants 109, Ill,
abscisic acid accumulation 49 113,119,133
DNA hybridization kinetics 251 nitrite reductase 108
Subject Index 285
photosynthetic mutants 63, 64 gene expression in pollen 249, 250

scald disease 242 genes:
storage proteins 209, 211, 215 Adh2253
temperature sensitive mutants 63 Aps2253
trisomics 225 bI 49
variety identification 222 Est4253
host-pathogen interaction 173 fIc 39,41-44
host specific toxin 240 ga(Ve-182/270/335) 4,5
HV-toxin 240, 241 gh43
Hyoscyamus muticus Got3253
auxotrophs 263, 265, 268-270, 272 Is 49
nitrate reductase mutants 109, 114, not 39,41,43
117 Pgil 253
temperature-sensitive auxin auxo- Pgm2253
troph 272 sit 39,41-43
hypernodulation 180 sI2 5
hypersensitive response 172 torosa-2 42
EPS mutants 174 wd 40
host-parasite interaction 235 yg6 4,5, 14
hypoxanthine auxotroph 268, 269 leaf epinasty 41
microgametophytic selection 253
infection thread 165 non-germinating mutants 17
invertaseless mutants 266 resistance to Alternaria alternata 240
Ipomoea batatas 5 Lycospersicon hirsutum
iron-sulphur proteins, deficient mutant cold tolerance 253
64 microgametophytic selection 253
isoleucine auxotrophs 268, 269 lysine auxotrophs 268, 269
lysine content screening 215
Klebsiella pneumoniae 164
Macroptilium atropurpureum 169
lab lab bean - see Dolichos maize - see Zea mays
Lathyrus odoratus 4 Malus domestica 49
dwarf mutants 14 manganese binding in photosystem II
leghemoglobins 176 62
leaf epinasty 41 Medicago sativa
leghemoglobin 166, 167, 176 cultivar variation 176
cDNA cloning 170 Fusarium resistance 242
nitrate, effect on expression 171 glutamate dehydrogenase in pollen
Leptosphaeria nodorum 242 252
leucine auxotrophs 268, 269 leghemoglobins 176
Linum usitatissimum 235 mutants 178
lipopolysaccharide (LPS) 173 nodulins 170
Lotium perenne 5 Medicago (barrel medic) 175
lupin ~ see LupinuspDlypnyllus megagllmetophytic gene expression
Lupin us polyphyllus 255,256
leghemoglobin 167, 176 Melampsora lini 233,235
nitrate sensitivity 175 metal tolerance 253
Lycopersicon esculentum 253 methionine auxotrophs 268, 269
abscisic acid mutants 39,45,46 methyl methane sulphonate 264
Adh cloning 82 mevalonic acid (MVA) 36
AL-toxin sensitivity 240, 241 microgametophyte
auxotrophs 260, 263 gene expression 254
286 Subject Index
selection 252, 253 chlorate resistance 109, 133

microspore selection 252 cnx gene function 126
millet - see Pennisetum americanum host-pathogen interaction 173
Mimulus guttatus Mo-co mutants 124, 125
metal tolerance 253 genes:
pollen selection 254 clr19 118
mitochondrial DNA cnxA/Al/A2/B 116,126-135,
homology with chloroplast DNA 69 137
MNNG - see N-methyl-N'-nitro-N- nial/2 111, 116, 118, 132-135,
nitrosoguanidine 141
molybdenum cofactor (Mo-co) 105 ff. nitrate assimilation mutants 116
mutants 123 ff. nitrate reductase deficient mutants
monosomic series 223 132-135
morphogen 194 somatic hybridisation 135
multilocus adaptations 251 streptomycin resistant mutant 135
mutagenesis in vitro 263 temperature-sensitive mutants 272
mutagens transformation 95
60CO gamma rays 268, 269 Nicotiana tomentosiformis 118
ethyl methane sulphonate, EMS 117, nicotinamide auxotrophs 268, 269
264 nitrate assimilation 101 ff.
N-ethyl-N-nitrosourea, ENU 117 mutants 112ff.
N -methyl-N' -nitro-N -nitrosoguani- nitrate reductase
dine, MNNG 117,268 diaphorase function 107
sodium azide 119,182 gene cloning 135 ff.
ultra violet irradiation 263 holoenzyme 104, 105, 107
X-ray irradiation 263 in vivo assay 119
mutant isolation in-vitro 260 molybdenum cofactor (Mo-co)
Mutator (Mu) 88, 89 105ff.
mutants 109, 111, 117, 123-126, 131,
ENU - see N-ethyl-N-nitrosourea 133,141,183,265,271
necrotrophic pathogens 234 NAD(P)H-NR 103
Neurospora crassa 178 NADH-NR 103,104
auxotrophs 265
nitrate uptake 183
Mo-co gene cloning 137, 138
molecular basis 102
Mo-co mutants 123
mutants 110
nitrate reductase 102 ff., 111
nitrate reductase mutants 131 nitrate-tolerant mutants 189
nitrite reductase mutants 131 nitrite reductase 108
Nicotiana plumbaginifolia mutants 130
auxotroph isolation 262, 268-270 nitrogen fixation 159 ff.
chlorate-resistant lines 125 non-fixation mutants 179
genes: variation in plants 174ff.
cnxA/B/C/LJ 115,126,127 nitrogenase 163
nia 115 nodulation 159ff.
nitrate reductase mutants 115, 118, autoregulation 190
124, 126, 127,266 development 161-168, 172
transformation 95 genes 171
Nicotiana sylvestris nitrogen fixation mutants 177
auxotroph isolation 262 phytohormones 191
abscisic acid sensitivity 46 nodulins 35, 159
Nicotiana tabacum cDNA cloning 170
auxotrophs 262, 268-270 nutritional mutants 260
Subject Index 287
oat - see Avena sativa G2-arrest 163

Oenothera nitrogen fixation 175
1sigma 65, 66 ureides 142
Renner complex 255 photoperiod response 18, 19
Oryza sativa photorespiration 56
abscisic acid accumulation 49 photosynthesis 55 ff.
dwarf mutants 7,10 apparatus 55, 56
endosperm proteins 207, 213 electron transport 58
genes: mutants 55-58,61,63,64
dx 2,4 photo system I, 58
dy 4,10, 11 mutants 64
gibberellin mutants 7, 10 photosystem II 58, 63
gibberellins 23, 25 components 61, 62
globulin 211 light-harvesting complex 61
glutenin 210, 211 QB membrane protein 62
nitrate reductase 103 Physalis minima 135
transient expression 94 Physcomitrella patens
oxygen diffusion barrier 165 auxotrophs 260
mutants 127
pantothenate auxotrophs 268, 269 phytoalexins
Parasponia 161 accumulation 172, 173
hemoglobin 169 infection response 239
bacteroids 169 Pisum sativum 161
Bradyrhizobium 167 Adh cloning 81
nodulation 168 apical senescence 18
pathogens endosperm protein 208
host resistance 233 flowering 18
mechanisms of plant defence 239 flowering genes 21
pea - see Pisum sativum G2-arrest 163
peanut - see Arachis hypogaea genes:
Penicillium chrysogenum dne18,20
Mo-co mutants 123 e 18
NR apoprotein gene 111 jix- 180
Pennisetum americanum hr/Hr 3, 18,20
drought resistance and abscisic acid la cry 3
48,49 la cry' 3, 16, 24
endosperm proteins 207, 212 Ie 2,3,6, 11-13, 16, 19,20,24,25
prolamins 211 Le 22
pentose phosphate pathway 57 if 18
peribacteroid space 167 Ih 3, 11,20
Petunia hybrida Ik 3, 16
auxotroph isolation 262 Is 3,11,20
nitrate assimilation mutants 116, 119 na 3,7,11,16,20,22-24
pollen competition 248 narJ/2 117, 130, 132
pollen storage tolerance 254 nod- 180
transformation 95 nodl/2/3 142
Pharbitis nil 4, 14 SnHr 3
phaseic acid (PA) 36 sn 18,20
Phaseolus vulgaris veg 18
chloroplast genome 67 gibberellins 23, 25
dwarf mutants 5,14 gibberellin mutants 7,8, 12, 14, 16
endosperm protein 208 leghemoglobin 167
288 Subject Index
nitrate-tolerant mutants 186 reductive pentose phosphate pathway

nitrate assimilation mutants 117, 120 57
nitrate sensitivity 175 Renner complex 255
nitrogenase mutants 181 replica plating 266
nodulins170 Rhizobium
non-nodulation mutants 178, 179 EPS-deficient mutant 174
Rhizobium 175,176 hormone synthesis 191
selective fertilization 255 hup system 163
slender phenotype 16 legume symbiosis 171
wilty mutants 39,41 nitrate inhibition of symbiosis 183
plant gene transfer systems 140 nodulation 172
plant-pathogen interaction 233 nodule types 168
Plantago insularis nodulin induction 170
thiamine auxotrophs 260 non-legumes 194
pollen NR deficient mutants 142
alcohol dehydrogenase activity 75 transposon mutagenesis 238
competition 247, 251 R. leguminosarum 176,180,181
gene expression 250, 254 R. meliloti 166, 174
genotypes 247 R. trifolii 166
pollen tube growth rate 248 Trifolium subterraneum 176
selection 249, 252 xanthine dehydrogenase 142
size 252 rhizobotoxin 177
storage tolerance 254 Rhyncosporium secalis 242
style interactions 255 ribulose-l,5-bisphosphate carboxylase
postmeiotic gene expression 249 57
precocious germination mutant, Oenothera 65
abscisic acid mutants 44 ribosome mutants 63, 66
Zea mays 46 rice - see Oryza sativa
prolamins 210, 211 Ricinus communis 5
gene families 217 Robertson's Mutator (Mu) 88, 89, 93,
gene evolution 211 136
protoplasts root exudate 185
mutagenesis 261 Rosa
transient expression 94 auxotroph enrichment 267
Prunus persica 5 nitrate reductase mutants 117, 265
pseudoinfections 165-173 rye - see Secale cereale
Pseudomonas
plant-bacterial interactions 173, 195 Scenedesmus
Pseudomonas solanacearum 173 water splitting system 61
Pseudomonas syringae 242 Secale cereale
avirulence gene 236 chloroplast ribosome deficient
blight of Glycine max 236 mutant 66
cosmid library 236 endosperm protein 207
trans po son mutagenesis 238 secalin genes 225
Puccinia coronata 240 storage protein evolution 211
Puccinia sorghi 239 temperature sensitive mutant 66
purine auxotrophs 270 seeds
pyruvate decarboxylase 78 dormancy 2
dormancy mutants 17,39,44
QB membrane protein 62 abscisic acid 35, 39, 44-46
selective fertilization 255
red clover - see Trifolium pratense Septoria nodorum 242
Subject Index 289
Sesbania 161 38,42

Silene armeria subterranean clover - see Trifolium
flowering mutants 21 subterraneum
genes: sucrose synthetase 78
ef 5 supemodulation 188
s 5 suppressor mutant 39
Silene dioica symbiotic nitrogen fixation 174 ff.
metal tolerance 253
pollen selection 254 T -DNA genes 234
Siratro temperature-sensitive mutants 271, 272
nitrate sensitivity 175 auxin auxotroph 272
transformation 194 70S ribosomes 63
sodium azide 119, 182 Secale cereale 66
Solanum nigrum embryogenesis 270
triazine resistance 62 threonine auxotrophs 268, 269
Solanum tuberosum thylakoid
abscisic acid mutants 39 electron transport processes 56
auxotroph isolation 263 deficient mutants 63
dr 39,41 light reactions 57
host-pathogen interaction 173 proteins 61
Phytophthera infestans resistance 173 Ti plasmid 137
Pseudomonas attachment 173 plant transformation 95
water stress 42, 44 tomato - see Lycopersicon esculentum
wilty 41 toxins in plant disease 239 ff.
somaclonal variation 87 Tradescantia paludosa 250
somatic hybridisation 134 trans-3-hexadecenoic acid deficient
nitrate reductase mutants 124 mutant 62
sorghum - see Sorghum bicolor transformation systems 140
Sorghum bicolor transient expression 94
drought resistance 48 transition polypeptides 76, 77
endosperm proteins 207, 212 transposable elements 238
grain protein fractionation 216 Adh mutants 76
prolamins 211 transposon mutagenesis 76, 89, 238
soybean - see Glycine max Trema 168
Sphaerocarpos donnelli triazine resistance 63
auxotrophs 260, 265 Trifolium incarnatum 178
spinach - see Spinacia oleracea Trifolium pratense
Spinacia oleracea dw15
chloroplast genome 60 mutants 5, 178
homology between organelle DNAs 69 nodulation, genetic variation 176
nitrate reductase 104, 107 Rhizobium 162
nitrite reductase 108 transformation 194
sporophytic gene expression 250 Trifolium subterraneum
squash - see Cucurbita leghemoglobins 176
stem nodulation 161 nitrate sensitivity 175
stomatal opening Rhizobium 176
abscisic acid 35 Tripsacum 94
abscisic acid mutants 40 Triticum
storage proteins pollen competition 248
messenger RNAs 208 Triticum aestivum
DNA sequencing 210 abscisic acid accumulation 49
stress induction and abscisic acid 36, bread quality 222
290 Subject Index
chlorate toxicity 109 xanthoxin 36

drought resistance and abscisic acid
48 Zeamays
endosperm proteins 207, 212, 214, abscisic acid 45, 49
217 abscisic acid-deficient mutants 39, 48
genes: alcohol dehydrogenase 74, 252
d324 Adh genes 75, 78, 79, 90
Gli-l/-AI/-BI 218,222,224,225 carotenoid biosynthesis 43
Glu-l/-AI/-Bl/-DI 218,221,224 carotenoid-deficient mutants 39, 42
Rhll/2/3 4,14,15,24,49 drought resistance and abscisic acid
gibberellin insensitive mutants 24 48
gliadins 210, 217 endosperm genotype 45
glume blotch 242 endosperm proteins 207, 209, 213
glutenins 210, 217, 221 genes:
high-protein genes 216 Adhl/2 74ff.,248
monosomics 223 Adhl-C" 88
nitrate reductase mutants 109 Adhl-FC" 87
nitrite reductase 108 aI39,43,239
resistance to Septoria 242 anI 3,8,9
storage proteins 208, 209, 211, bzl 238
223-225 dl/2/3/5/8 2,3,7-11,24,25
transient expression 94 jloury-2 215,219-221
variety identification 222 goll 250
Triticum monococcum Hm/2240
transformation 94 mi23
tryptophan auxotrophs 268, 269 nal/23
opaque-2/-6/-7 215,219,220
ultra-violet irradiation 263, 264, 268, pel 3
269 ps 39,43,45
uracil auxotrophs 268, 269 Rpl 239
ureides 142, 165,,168 vpl/2/5/7/8/9 39,41-43,45,48
w3 39,42,43
valine auxotrophs 268, 269 waxy 248
Victoria blight 240 wi 40
Vigna y939
nitrogen fixation 191 gibberellins 23, 24, 25
pollen competition 248 gibberellin mutants 7-9, 14, 16
Vigna sinensis high lysine 220
Bradyrhizobium test 175 high-protein genes 216
violaxanthin 36 mitochondrial DNA 69
vitamin auxotrophs 270 nitrate reductase 103
pollen 252, 254
water relations pollen-style interactions 255
abscisic acid 38, 39, 41, 42 prolamins 211
abscisic acid mutants 44-46 proline auxotroph 260
water splitting system 61, 63 protoplasts 94
wheat - see Triticum aestivum resistance to pathogens 239
white clover - see Trifolium pratense somaclonal variation 87
T-toxin sensitive lines 240,241
X-ray irradiation 263 transient expression 94
xanthine dehydrogenase 111-117, translocations 226
119, 121 ff. transposable elements 238, 239
Subject Index 291
trisomies 226
viviparous mutants 41, 46
zein 210, 220, 226
r~antgene
E. S. Dennis, B. Hohn, Th. Hohn (Managing Editor), P. J. King,
J. Schell, D. P. S. Verma
The first volume
Genes Involved in Microbe-Plant Interactions
Edited by D. P. S. Verma, Department of Biology, McGill University, Mon-
treal, Canada, and Th. Hohn, Friedrich Miescher-Institut, Basel, Switzerland.
1984. 54 figures. XIV, 393 pages. ISBN 3-211-81789-1
Contents:
Recognition: F. B. Dazzo and A. E. Gardiol: Host Specificity in Rhizobium-Legume
Interactions. - A. Matthysse: Interaction of Agrobacterium tumefaciens with the
Plant Cell Surface. - Symbiosis: D. P. S. Verma and K. Nadler: Legume-Rhizobium-
Symbiosis: Host's Point of View. - B. G. Rolfe and J. Shine: Rhizobium-Legumino-
sae Symbiosis: The Bacterial Point of View. - B. J. Miflin and J. Cullimore: Nitrogen
Assimilation in the Legume-Rhizobium Symbiosis: A Joint Endeavour. - N. J. Brewin:
Hydrogenase and Energy Efficiency in Nitrogen Fixing Symbionts. - A. Moiroud and
V. Gianinazzi-Pearson: Symbiotic Relationships in Actinorhizae. - V. Gianinazzi-
Pearson: Host-Fungus Specificity, Recognition and Compatibility in Mycorrhizae. -
R. P. Legocki and A. A. Szalay: Molecular Biology of Stem Nodulation. - Plant Tu-
mor Induction: J. Tempe, A. Petit, and S. K. Farrand: Induction of Cell Proliferation
by Agrobacterium tumefaciens and A. Rhizogenes: A Parasite's Point of View. -
J. Hille, A. Hoekema, P. Hooykaas, and R. Schilperoort: Gene Organization of the Ti-
Plasmid. - C. I. Kado: Phytohormone-Mediated Tumorigenesis by Plant Pathogenic
Bacteria. - Plant Pathogens and Defence Mechanisms: N. J. Panopoulos, J. D.
Walton, D. K. Willis: Genetic and Biochemical Basis of Virulence in Plant Pathogens.
C. A. Ryan: Defense Responses of Plants. - Subject Index.

~?ant[jene
E. S. Dennis, B. Hohn, Th. Hohn (Managing Editor), P. J. King,
J. Schell, D. P. S. Verma
The second volume

Genetic Flux in Plants
Edited by Barbara Hohn, Friedrich Miescher-Institut, Basel, Switzerland, and
Elisabeth S. Dennis, CSIRO Division of Plant Industry, Canberra City,
Australia.
1985. 40 figures. XII, 253 pages. ISBN 3-211-81809-X
This volume gathers together for the first time the most recent information on plant
genome instability. The plant genome can no longer be looked upon as a stable enti-
ty. Many examples of change and disorder in the genetic material have been report-
ed recently. Chloroplast DNA sequences have been found in nuclei and mitochon-
dria. Mitochondrial DNA molecules can switch between various forms by recombina-
tion processes. Stress on plants or on cells in culture can cause changes in chromo-
some organization. DNA can be inserted into the plant genome by transformation
with the Ti plasmid of Agrobacterium tumefaciens, and transposable elements pro-
duce insertions and deletions.
Contents:
Movement of Genetic Information from the Environment to the Plant: H. Fraenkel-
Conrat: Viruses. - G. Gheysen, P. Dhaese, M. Van Montagu, and J. Schell: DNA
Flux Across Genetic Barriers: The Crown Gall Phenomenon. - Movement of Genet-
ic Information Between the Plant Organelles: D. M. Lonsdale: Movement of Genet-
ic Material Between the Chloroplast and Mitochondrion in Higher Plants. - J. N.
Timmis and N. Steele Scott: Movement of Genetic Information Between the Chloro-
plast and Nucleus 61. - R. J. Kemble, S. Gabay-Laughnan and J. R. Laughnan: Move-
ment of Genetic Information Between Plant Organelles: Mitochondria-Nuclei. -
Movement of Genetic Information Within Plant Organelles: R. R. Sederoff and C. S.
Levings III: Supernumerary DNAs in Plant Mitochondria. - A. J. Bendich: Plant Mito-
chondrial DNA: Unusual Variation on a Common Theme. - R. B. Flavell: Repeated
Sequences and Genome Change. - C. A. Cullis: Sequence Variation and Stress. -
S. L. Dellaporta and P. S. Chomet: The Activation of Maize Controlling Elements. -
W. R. Scowcroft: Somaclonal Variation: The Myth of Clonal Uniformity. - Subject
Index.

Plant Gene Research

Încărcat de

Informații document

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Plant Gene Research

Încărcat de

Drepturi de autor:

Formate disponibile

Plant Gene Research

Basic Knowledge and Application

Springer-Verlag Wien New York

Springer-Verlag Wien New York

Dr. Patrick J. King

This work is subject to copyright.

Library of Congress Cataloging in Publication Data

Biologists ask how the growth, development and behaviour of organisms

Chapter 1 Gibberellin Mutants

Chapter 2 Genetic Aspects of Abscisic Acid

Chapter 3 Mutants as Tools for the Elucidation of Photosynthetic

Chapter 4 Maize Alcohol Dehydrogenase: A Molecular Perspective

A. Attempts at Expression in Cereal Tissue Culture Cells 94

Chapter 5 The Molecular Genetics of Higher Plant Nitrate Assimilation

Chapter 6 Plant Genetic Approaches to Symbiotic Nodulation and

IX. Conclusions 194

Chapter 7 Endosperm Proteins

Chapter 8 Molecular Approaches to Plant and Pathogen Genes

Chapter 9 Gametophytic Gene Expression

IX. Conclusions 255

Subject Index 281

Department of Botany, University of Tasmania, Hobart, 7001, Australia

An enormous amount has been written on the gibberellins (GAs) during

II. Selection and Identification of Mutants

The selection of mutants influencing GA metabolism or sensitivity has

available. However, selection based upon characters evident early in the

Species Gene Phenotype Proposed Gene Key Reference(s)

Zeamays d} dwarf blocks GA20 to GAl Spray et al., 1984

Species Gene Phenotype Proposed Gene Key Reference(s)

ga-2 non-germi- blocks GA synthesis Koornneef and

Species Gene Phenotype Proposed Gene Key Reference(s)

Ve-182 semi-dwarf partially blocks GA Zeevaart, 1984

Selection based upon changes in morphological characters can provide

GA-insensitive mutants t,d,.d'

Other limiting factors GA, (active) GAB (inactive)

An important impetus in determining the action of GA synthesis mutants

IV. Internode Length

cell-free system developed from young etiolated shoots of d5 and normal

dwarf gave a ratio of approximately 0.02 compared with a ratio of approxi-

elicit a biological response. In addition, GA 12 -aldehyde was not as effective

problems facing plant hormone research (Vanderhoef and Kosuge, 1984).

VI. Flowering and Senescence

nation Sn Hr was responsible for GA production in the leaves in SD which

of GA29 . Ingram (1980) examined the levels of endogenous GAs by both

VII. Site of Action of GA Mntants

The site of action of GA synthesis mutants has only been examined in a

since GAl is found in the highest concentrations in the young expanding

seeds requires critical examination especially in view of the large number

and show similar elongation. Unfortunately, these cultivars are unrelated

approach, involving genetics, anatomy, physiology and biochemistry is

I wish to thank Drs. T. J. Ingram, I. C. Murfet, W. C. Potts and J. J. Ross

Allan, R. E., Vogel, O. A, Craddock, J. C., 1959: Comparative response to gibber-

Proano, V. A., Greene, G. L., 1968: Endogenous gibberellins of a radiation induced

Smith, R. R., 1974: Inheritance of a gibberellin-responsive dwarf mutant in red

Genetic Aspects of Abscisic Acid

Department of Genetics, Agricultural University, 53, Generaal Foulkesweg,

A. Abscisic Acid as a Plant Hormone

B. Biosynthesis and Metabolism ofABA

C. Genetic Aspects ofABA

C02H Phaseic acid (PA)

C02 H Di hyd rophaseic acid (DPA)

concentrates on the use of genetic variation for ABA in basic plant

Table 1. Predicted Phenotypes of ABA Mutants

Effects compared to wild type